Hacker News Reader

Comment by cr4zy | Hacker News Reader

Similar question on Quora from 2012[1] per sense:

  Vision: 10Mb/s => 100Mb/s
  Hearing: 30Mb/s
  Touch: 135Mb/s
  Smell: 100k neurons
  Taste: 100kb/s
  Proprioception: ??
  Balance: ??

  Total: ~10Mb/s=>~1Gb/s

Internal brain bandwidth is also worth mentioning as this is the last remaining wetware advantage over hardware due to the three dimensionally fully connected heterogeneous cortical substrate. I can't seem to find a figure on that though.

https://www.quora.com/How-much-bandwidth-does-each-human-sen...

View on HN

I really question the hearing one, at least.

Shouldn't this basically just be the bandwidth of a headphone signal which is at the lowest quality where you can hear degradation if it goes any lower?

For me that's something like 3MB per minute, per well-compressed ogg files, or 50KB per second. Yet that answer says my hearing bandwidth is actually 600x higher than that.

I can imagine some small difference, but to say my ear has evolved to transmit 600x more sound data than I can actually perceive sounds off.

> Shouldn't this basically just be the bandwidth of a headphone signal

One issue with this approach is that it's very hard to tell the difference between data that was never sent versus data that is there but our developing brains were trained to ignore because it didn't turn out to be useful.

A simple example of this would be language differences, where certain important language features simply aren't noticeable to non-native speakers, despite having the same sensor-hardware.

Well, don't neurons get pruned as you grow up? So you don't necessarily have the same hardware.

I was thinking more about everything leading up to the brain, such as the composition of your retina and optic nerve.

I'm wondering... For the most efficient lossless audio format in existence now, is it still inefficient with respect to what a human can actually hear? For example, can it store frequencies above what a human can hear (20 kHz)? Are there opportunities to make it more efficient, assuming that we only care about humans? Can a FLAC file store a sound that only a dog could hear? If so, maybe the bandwidth is even lower than 50KB/s.

Or going in the opposite direction - do our hearing neurons transmit data that the brain throws away? If so, then the bandwidth goes higher than 50KB/s.

Lossless encodings don't throw anything away that was present in the original source file; if you start with a recording chain that can store 200,000 samples per second from an adequate microphone, you can resolve 100KHz - epsilon frequencies. Typically, though, the source files that pass through lossless encoding have previously been limited to 22KHz.

Lossy encodings, on the other hand, all work on the basis of a model of perceptibility that throws away information deemed less perceptible.

You can hear frequencies much higher than normal by pressing the sound source against the cranium.

Reply on news.ycombinator.com