Hacker News Reader

Comment by pasta | Hacker News Reader

Once a producer told me that they mastered at 16bit 44.1kHz and then pressed records and CDs from those masters.

Ofcourse the records sold better because they sound much better...

Of course the records will sound better; they will have their dynamic range squashed to fit the medium. Compression sounds nice: all the quiet parts are so clear, yet the transients are cleanly articulated.

Compressed audio, like from a record player, also works better with a cheap back end. Protecting the record player's needle from skipping out also helps the amplifier and speakers.

Don't confuse the sound of compression with the sound of music being louder. Compression can help with poor quality equipment or noisy listening environments (e.g. cars), but it also adds audible distortion and is tiring to listen to for long. Most dynamic range compression in music is there to trick people into playing it louder (see "loudness war"), and louder music sounds better. Of course, you can get the same benefit without the distortion by turning the volume up yourself.

But despite vinyl being poorly suited to high dynamic range, the vinyl release sometimes has higher dynamic range than the CD/download release. This is probably because vinyl purchasers are thought to be more discerning customers. This isn't always the case; the most popular dynamic range measurement algorithm (see http://dr.loudness-war.info/ ) is fooled by all-pass filtering, and vinyl record players do this, so the measurements will be falsely high for vinyl. Often the supposedly higher dynamic range vinyl release is actually from an identical master.

> Most dynamic range compression in music is there to trick people into playing it louder

That's a relatively new phenomenon, together with the distortion.

Compression doesn't add audible distortion, except if it is used in an extreme way to boost the signal so that it sits at close to maximum volume.

Compression is a tool that can combat distortion. Reduce the transients; reduce the clipping.

Compression can be used to obtain more consistent distortion on purpose over longer envelopes of the signal, like in a common configuration in guitar audio: compressor followed by distortion effect.

Obtaining more consistent distortion can mean consistently less of it, not only consistently more.

Audiophile grade equipment doesn't clip transients at normal listening volume. I agree that compression can help with poor quality equipment, but at the cost of permanently harming the sound on good equipment. Nobody complains about acoustic live music having too much dynamic range. The proper place for dynamic range compression is in the playback equipment, not baked into the audio data (DVDs/Blurays get this right with AC3 DRC). And compression used as an effect on individual instruments isn't what "loudness war" is referring to.

What's the dynamic range of vinyl? Usually 70 dB, whereas 16bit PCM is around 90 dB. That's what OP was probably referring to. Since the master was done for the CD then compressed for the vinyl it sounds softer/cleaner.

CDs do have superior dynamic range, but when modern mastering squashes everything within a few dB of full scale it hardly matters. But I disagree that vinyl releases[0] typically have lower dynamic range than CD releases (most often it's the same, sometimes it's higher, and only very rarely it's lower), and I disagree that lower dynamic range sounds better on good equipment.

[0] Talking about popular music only. I don't know enough about classical to say how it's done there.

I have no actual data about the musical signals present on the discs themselves (be they made of any kind of plastic with grooves or pits), I just mentioned the technological limitations of them. ( see also https://wiki.hydrogenaud.io/index.php?title=Myths_(Vinyl)#My... )

However, as far as I know the loudness wars is coming to an end finally: https://en.wikipedia.org/wiki/Loudness_war#2010s "Analysis suggests that the loudness trend may have peaked around 2005 and subsequently reduced, with a pronounced increase in overall and minimum album DR (crest factor) for albums since 2005."

The proper place for dynamic range compression is basically upstream of anything that requires it or benefits from it.

In the case of a phonograph player (which unconditionally qualifies as "poor equipment"), we have to prevent the needle from jumping out of the groove. This is downstream from the record itself, and so the signal that is on the record is manipulated.

Record pressers do have very good DA-converters. That's what they do!

It doesn't matter how good your DAC is if your signal path includes scratching the sound waves into lacquer, stamping it into vinyl and scraping it with a stylus. Compared to digital distribution formats, it's utterly Rube Goldberg-esque.

  Performance
  Microphone
  ADC
  Processing
  Mixing & mastering
  DAC
  \
   \
    \
     Scratched into lacquer
     Electroplating
     Stamping into vinyl
     Scraped with a stylus
     Phono amplifier
    /
   /
  /
  Amplifier
  Speakers
  Room
  Ears

Well, all I'm saying is that it could... If you have a really shitty DAC for some CD player, or you have a really good DAC to convert it to record, it could be the DAC quality that creates a superior sound. All other factors always reduce quality; only the DAC may actually explain the improved quality. That's my point. Which I made jokingly.

Yes you are right, vinyl might sometimes be better than horribly, catastrophically broken digital.

In reality, bad DACs are uncommon, mostly found in disposable-grade consumer products and on computer motherboards. (And even then the problem is poor implementation, not a poor DAC chip per se.) Anything sold into the mainstream Hi-Fi or audiophile market is nominally perfect relative to our ears’ potential.

More fundamentally I think you are overestimating how bad a DAC can be. The “bad DAC” is mostly a myth invented by the manufacturers of expensive DACs and their gullible customers. It’s a scam and far too many people have fallen for it.

(Edit: To be more precise, these DACs aren’t scams, there is legit engineering going on, but they are pointless if your intent is listening to the output with human ears rather than test equipment.)

Reply on news.ycombinator.com