4 ms·

I'm fairly certain he is speaking of data compression in the article, rather than dynamic volume compression. The problem with sending compressed signals is th

• uxp100 10y ago
  No, I don't think so. They specifically talk about adding compression only into their newscast unit. Why would they data compress only one segment? Also, Compression sounds good. Too much, or badly configured compression sounds bad, obviously. But don't throw it all away. (Not that producers are at any risk of doing so.)
  • jws 10y ago
    I was thinking of remote newscast, where you might need to fit your data reliably over a cellular link, but now that I think about it, I don't think NPR does much remote newscast.
    • joombaga 10y ago
      Are you conflating dynamic range compression with data compression?
      • colejohnson66 10y ago
        Well, technically speaking, dynamic range compression /does/ make an MP3 smaller (IIRC)
        • riprowan 10y ago
          So yes, you are conflating these things.
          • colejohnson66 10y ago
            > you But I'm not OP?
            • riprowan 10y ago
              Fair enough.
        • joombaga 10y ago
          Right, but is it done for that purpose?
          • colejohnson66 10y ago
            I don't know how MP3 works, but I was under the impression that a lower bitrate worked by compressing the dynamic range (or at least that was a side effect)
• OliverJones 10y ago
  No way he's talking about data compression, with respect. He's talking about frequency response adjustment and range compression (sometimes called "companding"). A while ago NPR used to use 2B ISDN (112Kbps in the US) for remote locations. I believe they used G.722 codecs for low-latency applications (conversations) and MPEG layer II for higher quality stuff. That was data compression. But the NPR distribution system isn't so bandwidth limited as to require crushing the data.
• chiph 10y ago
  No, he's talking about audio (sound) compression. https://en.wikipedia.org/wiki/Loudness_war https://en.wikipedia.org/wiki/Loudness_war This is one reason why I've stopped buying "Remastered" CDs - they almost always sound worse. Shop at Discogs.com where you can specify the release date, or buy from a quality-oriented label like mofi.com
  • colejohnson66 10y ago
    The key words are "almost always". While I haven't checked, I heard that Pink Floyd's remaster of "The Wall" and some of Led Zeplin's sound better because they focused on removing noise instead of making it louder.
    • chiph 10y ago
      > The Wall Interestingly - I have that CD (2-disc set) from way back in 1984 when it was first released (on the Harvest label in Europe), and it sounds like rubbish. In the early days of CDs the labels were dumping any title they thought would sell onto poly-carbonate, and didn't care much about cleaning it up. There were stories of CDs that still had the RIAA equalization curve (from LPs) still on them. I'd love to get the Mobile Fidelity pressing, but they were a limited release and is now selling for over $200.
      • enneff 10y ago
        I think the bad mastering of many early CDs is mostly because mastering engineers didn't know how to master CDs. They were still employing the same techniques they'd been using for vinyl, where they were hardly applicable. (I think that's where the RIAA curve thing comes from, which is unlikely to be literally true as the RIAA curve would remove almost all the bass from the recordings.)