7 ms·
I just tested the built in speakers and mic on my 3 year old MacBook Pro using sine waves and a spectrum analyser. It's perfectly capable of sending and pickin
by elektronaut 13y ago
I just tested the built in speakers and mic on my 3 year old MacBook Pro using sine waves and a spectrum analyser.
It's perfectly capable of sending and picking up a 20kHz signal by itself. The frequency response starts dropping off drastically at around 14kHz. Around 21kHz seems to be the practical limit.
However, there seems to be a bit of distortion going on. I'm not sure it's possible to send a clean inaudible signal without introducing lower frequency components.
- logn 13y agoI've worked in computer music/audio for years. Most people and PCs can reach 20kHz. It's no coincidence, human hearing drops off around that point and thus the 44.1kHz sample rate was chosen to match that (you can transmit frequencies at max of 1/2 the sample rate, thus 22.05kHz). But modern sound cards can reach 96 or even 192kHz sample rate, able to transmit ~50-100kKHz tones (other factors aside). Further, modern sound cards can have amplitudes expressed as 24-bit numbers, providing very high fidelity/accuracy. So, you may be able to hear higher than 20kHz. Even if you increase your sound card sample rate, best test would be to try something analog to eliminate other bottlenecks in the audio processing. But you can easily tweak your audio settings and generate sine waves with this program: http://audacity.sourceforge.net/ http://audacity.sourceforge.net/
- robryk 13y agoQuite a few people can hear 20kHz. I can still clearly hear 18.5kHz (or what my laptop's speakers emit when I feed a 18.5kHz sine) and 19-20kHz gives me headaches.
- S_A_P 13y agoOk so maybe it's possible to send audio and receive it via a MacBook Pro. However, how much data could that low bandwidth signal carry? I'm even more curious how randomly transmitting "evil bits" this way could spontaneously infect a PC at the lowest levels. Something just doesn't add up here.
- nknighthb 13y agoAbsolutely no one is suggesting that audio is being used to spontaneously infect previously uninfected systems. No one has ever suggested that except for people who have failed to carefully and completely read the article. The claim is infected machines are using it to communicate, not that uninfected machines are being infected by it.