Kim Lajoie's blog

More things you can’t hear

by Kim Lajoie on March 7, 2012

Justin Colletti:

Compared to the much of the animal kingdom, human beings have pretty terrible hearing. We have poor powers of echolocation, especially for sounds that come from behind us, we can only hear a relatively narrow bandwith of 20Hz-20kHz, and we’re easily fooled by illusions.


As the votes came in, the crowd was split at first, and then began to veer in favor of the software plug-in. Not only did a small-but-significant majority of listeners show a preference for the sound of the plug-in, they also believed that they had selected the hardware version, because they believed the hardware version should sound better.

There’s so much good stuff in there I could quote the whole thing. Go read it. Now.

Elliott Fienberg recently asked me on Twitter what bitrate I encode at. I use LAME MP3 @ 192kpbs CBR. I use 192kbps because that’s the point at which I can’t hear the encoding. I used to encode at 128kpbs back in the day, but started to hear the data loss at the top of the frequency spectrum.

And don’t get me started on dither.

I think, ultimately, it’s not practical to approach hearing (audio reception) and listening (audio interpretation). Our ears and brains are not microphones – psychology plays an inseparable part in our understanding of sound and music.

I also think that the quest in some circles for high-resolution audio as an endpoint format is misguided. CD resolution (notice I didn’t say ‘quality’) has a dynamic range that approaches 96dB and a frequency range that approaches 22.05kHz. We can argue all day about the quality drop-off at the extreme edges, but ultimately the music that we love fits within these bounds. That’s why Dynamic Range Day isn’t about higher-resolution formats. It’s about making better use of the resolution we’ve had available for the last thirty years.

What do you think?



6 thoughts on “More things you can’t hear

  1. Kapitano says:

    What do you think?

    I think…that when I tried experimenting with higher resolutions and bitdepths (88.2kHz, 24bit), that I could hear a small difference. But it was on the same level as the difference I could hear when fine-tuning an EQ or compression knob…before realising the box was bypassed and the difference was all in my head. We’ve all been there.

    So even if the difference is real, it makes a bigger difference to wear a different set of headphones, or listen in a different mood.


    If I can follow this up with a related question: Equal temperament vs Just temperament. If you don’t know which you’re listening to, does it make a significant difference?

  2. Kim Lajoie says:

    I’ve tested working at 96kHz a while ago. I could hear a difference, but I decided that it wasn’t worth the the 2X processing load. Most decent plugins are oversampled where it counts anyway. Nowadays I could probably do some projects at 96kHz, but the added cognitive load of having to remember what samplerate each project is at and having to deal with multiple samplerates isn’t worth the headache. And sometimes I don’t know ahead of time whether a project will be low in CPU load (thus allowing me to record at higher samplerate). I’d hate to start working at a high samplerate only to run out of CPU halfway through. That’s a real workflow killer.

    Listening in a different mood is a good point. I think I’m pretty objective in my listening, but I’d have no trouble believing that most people perceive sound and music differently depending on their mood. I wonder if anyone’s done a study on it? It’d be interesting to know. We could consider people’s mood as we compose or mix – (How would this sound to an angry person? How would this sound to a depressed person? How would this sound to a joyous person?).

    I think the equal-temperament vs just-intonation question is like any subtle difference. B or Bb tonic? Real analogue synth or digital emulation? Tele or strat? SSL or API? These are choices that, on their own, make hardly any difference to the listener (they’re usually listening to the lyrics, the photography and their own assumptions anyway). But the composers, producers and engineers are responsible for making these minor decisions. And these minor decisions *must* be made. If the approach is haphazard, the end result will be unfocussed or creatively washed out. If each of these minor decisions are made deliberately in the context of the overall creative direction, the end result will have a focus and cohesion that borders on magical.

    Also – you’ve got an interesting blog!


  3. notwa says:

    I’d use 192kbit/s ABR instead of CBR, as CBR is always somewhat less efficient.

  4. Kim Lajoie says:

    I know VBR produces better results than CBR. The reasons I haven’t switched to VBR (yet) are:
    1) Back in the day, some MP3 players had trouble with VBR files. CBR has never given me problems. Yes, I know that in 2012 it’s unlikely that any new MP3-playback devices balk at VBR, but there are lot of old and weird devices still floating around. It’s the same reason I don’t use AAC – compatibility.
    2) If CBR @ 192kbps is sufficient audio quality (to me), than the only benefit of going to VBR would be to reduce the file size. If (for example) CBR @ 192kbps is equivalent to VBR @ 128kbps, I’d only save a few MB per file. In other words, the size difference would be unnoticeable. And not worth the (admittedly small) compatibility risk. In fact, the marginal file size reduction isn’t worth anything to me.

    Ultimately though, I only use MP3s for convenience – client/artist previews and freebee giveaways (i.e. email signup bait). I use uncompressed CD-resolution audio anywhere that matters.


  5. notwa says:

    As long as you understand the difference, it’s fine. I just can’t stand people that don’t have a clue how to encode and make nerds like me scream. On a side note, 192 is the bitrate I use as well. I can’t even hear the difference between 160 and anything higher on my iPod. Though if I’m encoding something for friends, I’ll use V0 as 192 isn’t always enough for everyone.

  6. Pingback: Audio perception and ABX testing | Kim Lajoie's blog

Leave a Reply

Your email address will not be published. Required fields are marked *