I recall an old physics text question asking why is it that only the treble sounds can be heard at the needle when playing a record? This is especially noted if (a)the power is off, and (b) one is merely twirling the turntable with the needle riding atop the record.
It must be that the needle is only capable of reproducing sounds of short wave length. But, why should any sound be produced at the needle at all?
It’s primarily because of the way the audio on phonograph records was compressed. In order to fit as much audio as possible on the record, the bass audio was compressed–that is, reduced in volume. This was compensated for in the playback electronics, and amplifiers had a “phono” input just for this purpose. If you plugged a phonograph turntable into a normal audio amplifier, the resulting sound would be tinny as well. The reason for all this trouble is because of the way the ear hears sounds. All else being equal, lower pitched sounds sound quieter than high-pitched sounds of the same amplitude, because higher frequencies carry more energy.
Isn’t it sort of like running your fingernail down the blackboard? The needle and its mount vibrate in response to the groove. This mechanical vibration sets a little bit of air in motion and you can hear faint, high pitched sounds. Sounds like someone inhaling helium-oxygen mix and then talking.
And they are high pitched, as QED said, partly because the ear is more sensitive to high frequency than to low. According to my old phyics text, it takes about 200 times more pressure in the sound wave at 27 Hz than at 14 kHz in order for a sound to be audible.
That part’s not quite right. The human ear is tuned to be more sensitive to a certain range of frequencies and at equal amplitude frequencies within that range will be perceived as louder. If you sweep a sine tone from 20 Hz to, say 20,000 Hz at a constant amplitude, you’ll effectively hear it as a gradual fade in and fade out, even though a db meter’s needle won’t budge.
This particular sensitive range corresponds with the bulk of the spectrum of the human voice as well as many “important” sounds - at least from a survival point of view.
More to the point of the OP, you should also consider that when you play a record with the amp turned off, the sound you hear is an artifact - it’s not what’s being amplified. What’s coming into your speakers is an electric current modified by the movement of the needle. The faint sound you hear with the amp off is that of sound waves generated by the needle itself. A small needle will not be able to generate sound waves with long wavelengths (i.e. low frequencies). It’s the same reason that woofers have large cones and tweeters have small ones.
This pinched from a posting (not mine) on another board:
About the “RIAA Curve”
It sounds tinny because it is. To get a flat frequency response you need to pipe the signal though a phono pre-amp. That’s why phono inputs aren’t interchangable with line-in sockets. It’s not just a level or impedance mismatch.
IIRC the RIAA eq-curve is pretty drastic ::googles for RIAA graph::
. . .
::comes back empty handed:: Can anyone do better?
That’s basically what I already said. Only it’s not a noise reduction method, it’s a compression method. If it weren’t done, the grooves would have to be spaced farther apart to accomodate the bass; this would obviously cut back on the amount of playback time available.
Actually I knew that, it’s just that the forum that I sto^H^H^H took the quote from was a debate about noise-reduction so the guy described the RIIA curve as such. Boosting/filtering treble I’d regard as noise reduction, cutting bass is to allow for longer playing times. My lazyness in not writing my own explanation now means I have to follow up the post 
Also “compression” I’d associate with DBX and Dolby rather than RIAA (which is just eq). Not saying you’re wrong just that you didn’t explain it the way I would, and did.
I have the vaguest recollection that the RIIA curve isn’t arbitary but depends on the physics of the cutting head. But I’ll do some research before saying anything definite.
There’s a graph of the RIAA recording and playback curves at the bottom of this page. They are fairly linear, but quite steep, as Small Clanger suggested.
Here’s some more information regarding the Fletcher-Munson effect. One of my first stereo amplifiers was a tube-type Pilot, and had a well written manual. The explanation for use of the ‘Loudness’ switch contained a paragraph on these two fellows.
RIAA curve notwithstanding, if the grooves were perfectly proportional, you would still only hear high pitched sounds from the needle. Its small cross sectional area makes it very inefficient at generating audible low frequency sounds. A small tweeter can produce loud high frequency sounds, but cannot produce loud low frequencies.
The needle is vibrating just like a speaker cone, it’s just an itty bitty speaker cone, so all your lower frequencies are beyond its abilities.