Today’s mailbag article, http://www.straightdope.com/mailbag/mrecordvoice.html , addresses the issue of why the way you hear your voice inside your own head sounds different from the way others hear your voice (bone resonance, etc.).
But when it comes to recording your voice, there’s another reason why your voice might sound “thinner”. And that is: most recording equipment sucks on toast. Unless you’re using a professional recording studio and playing your voice back on their equipment, you’re not going to get all the highs and lows of your voice reproduced faithfully. The cheapo Radio Shack[TM] tape recorder you got for your 10th birthday has about zilcho response in the bass range and probably has pretty poor response to the higher frequencies. And recordings of yourself on a telephone answering machine will be even worse: telephones have a built-in artificial cutoff at 3000 Hz on the high end and (I believe) around 30 Hz on the low end.
Quick-N-Dirty Aviation: Trading altitude for airspeed since 1992.
holding a horn type device or cupping a hand over the ear and mouth helps to direct the sound back to the ear and helps you to control the way your voice sounds…It seems to me one of the Bee Gees did this to get the tone he wanted from his voice.
The high end is definitely a problem, and the reason that listening to music over the phone almost never gets the desired response. “Hey, isn’t this song great?” “Well, actually, it sounds awful.”
But telephones don’t actually have absolute “cutoffs.” They have about a 3000Hz bandwidth, performing well between several hundred Hertz and around 3500Hz. Below several hundred the response drops quickly; same story above 3500Hz. By the time you get up into the 5000Hz or 6000Hz range, the volume is going to be off by 50-100dB, and therefore close to if not completely inaudible. Same story going down, although it’s not as much of a problem: a triple-low A is 55Hz, and very few people can speak or sing that low. Most male ranges go down to somewhere around 100Hz, with women bottoming out substantially higher.
While it is true that some recording equipment has less than faithful fidelity, that would not impact on the issue presented in the original question, which was the discrepancy between what you hear when you speak/sing, and what others hear, which you only hear when you listen to your recorded voice. This discrepancy exists regardless of the fidelity of the recording. Note please that the original question stated that the recorded sound sounded like what listeners heard from the speaker in person.
You’re right on the lack of hard cutoff on the phone. That’s the idea of ‘pink’ noise that they used in ‘blue boxing’. Mask a control tone with a burst of high-frequency static. The static gets filtered out a little in each hop, thus the control tone gets recognized more easily by the distant system. Drop the volume on the static until the last system in the line picks up the control tone.
But, that was with analog systems.
Now that most of us are on digital phone systems (analog copper from you to the telco, digital between telcos) is there a filter in there, or does the limit simply come in because of the sampling frequency?
A hard limit has quality problems, but then when sounds pass the nyquist (?) frequency (half the sampling frequency, the highest tone to be accurately camptured) aliasing problems seem like they’d cause more distortion.
Having a softer limit would make modems work better in analog systems, but where you need to count on accurate A->D conversion, a hard limit before the sampling step would provide better results.
nope, this effect is due to the dj positioning his mouth very close to the mic.
price has nothing to do with it. well, it does… but not the way you’re thinking. radio djs often opt for cheaper mics, because they heighten the ‘divine voice’ effect.
what is essential is invisible to the eye -the fox
This is known as the “proximity effect”. It is present in cardiod (and hypercardiod) microphones, it is not present in omnidirectional mics. As a rough rule of thumb you gain up to 6dB of “bass” response (<125Hz) as the sound source moves inside 6 in. from the mic.
I don’t know why any radio DJ would opt for a cheaper (and probably less responsive) mic, when any cardiod mic (regardless of price) will exhibit this effect.
I am unsure if, as a general rule, a cheap mic exhibits a greater proximity effect than a quality mic.
I know no one at the radio station my pop DJs for goes shopping for cheap mics.
The primitive equipment used at the time the telephone system was being constructed would introduce a significant amount of white noise, therefore the telephone companies decided at that time to limit the range of transmission to within the range of the average human voice - higher than the low and lower than the high. The switchboard was designed with such filters in place and we are now left with legacy technology that will only pass extended frequencies along. This is part of the reason why telephone modems cannot operate with the data through put of cable modems which have no such frequency filter (not to mention amplitude modulation vs. phase modulation.)