The Criterion LaserDisc sounds excellent. Although the mono tracks on the blu-rays certainly aren't bad by any objective standard, they're unremarkable next to the Criterion - their bass below 100 Hz is rolled off, as well as their high frequencies. This is most noticeable in scenes with music and of course the prom scene, which sounds absolutely thunderous on the LaserDisc.
The mono on every edition released since the SE DVD sounds basically the same. The Shout Factory and Arrow mono tracks are lossless only on paper; they're clearly transcodes of a lossy source (possibly the AC-3 track on the MGM blu-ray, if that's what Shout and Arrow were given by MGM), as evidenced by their sharp 20.5 kHz cut-off (partially obscured on the Arrow by dither noise). I'll stress that they sound worse than the LD because of their mastering, not their encoding.
what exactly is dither noise?
ReplyDeleteYou can find more accurate and eloquent explanations around the Internet, but it's essentially low-level noise applied intentionally to minimise additional errors when saving something at a particular bitdepth. It's typically a good thing to add when going from 24 bits to 16, but much less necessary when exporting to 24. The most commonly used distribution of noise is a TPDF dither, which some argue is too audible. In reality, unless you're crankling your volume to obscene levels and listening in a soundproof chamber, it's very unlikely you'll ever hear it.
DeleteThis is the Criterion LD of Carrie without any dither noise: https://imgur.com/ky9X91z
This is the Criterion LD of Carrie with an added TPDF dither: https://imgur.com/9ttyqlL
This is the Criterion LD of Carrie with iZotope's proprietary MBIT+ dither, which pushes the added noise to higher frequencies to make it theoretically less audible: https://imgur.com/HEUExOF
To me, none of this is anything to be particularly concerned about and falls well outside the objectives of this website, which are to explore *meaningful* differences in sound quality arising from mastering choices. It's less important than lossy vs. lossless compression, which I contend is already negligible relative to noise reduction and questionable EQ.
thanks
Delete