Wikipedia:Reference desk/Archives/Computing/2016 January 22

= January 22 =

How can the Internet send messages so fast?
Does it take longer to receive an e-mail from Beijing to New York City than it is to receive an e-mail from London to Paris? What affects the speed at which the messages are sent? What is actually connecting the electronic devices? Where are the Wi-Fi signals and telephone signals and radio signals coming from? Are non-human living things ever detrimentally affected by the artificially created signals? 140.254.70.165 (talk) 12:59, 22 January 2016 (UTC)


 * 1) Yes.
 * 2) Lots of things. One is, of course, the distance the messages travel, which you brought up. Other things include network congestion, how quickly every computer along the way processes the messages, and the speed of the transmission mechanisms (wireless, fiber optics, etc.) used.
 * 3) A series of tubes. Okay, more seriously, wires and fiber optic cables. See Internet backbone. Wireless networking is of course a thing, but because of less reliability and speed, it's generally only used for connecting users' devices (cell phones, laptops, etc.) to access points. The Internet backbone is all wired.
 * 4) Antennas.
 * 5) In general, no. The amount of energy deposited in objects by radio waves drops off rapidly the further you are from the source: the inverse-square law in action. And no organisms, as far as I'm aware, use radio waves for communication, so there's no interference issue. --71.119.131.184 (talk) 13:35, 22 January 2016 (UTC)


 * As a follow-up, to directly answer the question in the section title: computers are fast. Modern computers can have clock speeds over 3 gigahertz, meaning very roughly that they "do stuff" 3,000,000,000 times a second. (There are factors other than clock speed that determine the actual speed of computers, but this is fine as a very loose approximation for getting the point across.) Transmitted messages travel at the speed of light, for photons, or the slower-but-still-quite-fast-to-humans speed of electricity, for electrical signals in wires. --71.119.131.184 (talk) 22:39, 22 January 2016 (UTC)


 * You might be interested in reading a bit about older precursors to modern email, like FidoNet's FidoMail. It's not what we use now, but it is in some ways easier to understand. There, the goal was to link local BBSs into a network that could send mail internationally. Mail was relayed the distance of a local phone call at each step, and often steps would wait until night time when more phone lines were open. So a mail from NYC to Beijing might have taken 5 days, while Paris to London might have taken 2. SemanticMantis (talk) 16:34, 22 January 2016 (UTC)


 * The first time I got a reply to Internet mail within minutes, I was startled because I supposed that it took hours (not days!) to relay mail from California to New York. —Tamfang (talk) 05:19, 23 January 2016 (UTC)


 * Those of us whose long-distance email connections were by UUCP when the Internet wasn't available to the general public remember when it often did. You had to use email addresses with an explicit routing, in the style  /, and each of those hosts might only connect once an hour to the next one, or might not connect at all during business hours so as to save on long-distance telephone charges.  --76.69.45.64 (talk) 19:27, 23 January 2016 (UTC)
 * Back in those days (and, yes, I was around then too) - it was not uncommon for email to be slower than paper mail. As you say, many servers would save up email and make a phone connection to the next server just once or twice a day - and in a chain of servers, it might be that A would call B just after B called C - and the message would be held up for a couple of days as a result.  This was especially true when making international connections.  But then, sometimes, your mail would get there in under an hour.  Knowing when each server dialled out, combined with the ability to string together routes explicitly with the "bang" notation, enabled you to choose a specific path that you'd know would be fast at specific times of day. SteveBaker (talk) 15:39, 25 January 2016 (UTC)

Question about webcams with microphones
Is it possible to get a webcam with built-in mic, and that mic is good enough to hold its own as a microphone without using the webcam part? I want to record myself gaming with the webcam but I also occasionally want to do just voice-overs. Also, any suggestions for good webcam/mics for under 100 dollars? 2605:6000:EDC9:7B00:E017:92A9:4BB5:AD1D (talk) 19:20, 22 January 2016 (UTC)
 * No, that is not possible. I recommend buying a separate mic. Built-in mics are worthless. A good webcam is the Logitech C920 (or C910). Newegg sells C920s for $63.79 with free shipping, over at Amazon you pay 65 bucks. The Quixotic Potato (talk) 22:52, 22 January 2016 (UTC)
 * Thanks! I took your advice and bought it, also got a Blue mic to go with it, stayed within my budget of 100 dollars. 2605:6000:EDC9:7B00:6CCA:6B9A:F4EB:2251 (talk) 01:45, 23 January 2016 (UTC)
 * YVW. The Logitech C920 (and C910) are popular choices among YouTube/Twitch streamers. Many built-in mics in webcams are barely good enough for Skype, I cannot recommend them to anyone! The Quixotic Potato (talk) 03:03, 23 January 2016 (UTC)


 * This is about 5 years old information... I use to mess with cheap computer cameras. What I found what that the microphones were omnidirectional. So, they picked up every noise in the area... The fan in the computer, your tapping on your keyboard, the bird squawking outside the window... It doesn't lend itself to picking up only your voice. There were exceptions. I had a very nice panning security camera with a directional mic. You could point it at a person and pick up the person's voice very well - even if the person was about 25 feet from the camera. It even had infrared and starlight mode so you could use it at night. However, that camera was over $1000. 199.15.144.250 (talk) 20:38, 25 January 2016 (UTC)