4 ms·

Sending light across interstellar distances is not a trivial affair. No light source is perfectly collimated; per wikipedia [1] "The divergence of high-quality

Sending light across interstellar distances is not a trivial affair. No light source is perfectly collimated; per wikipedia [1] "The divergence of high-quality laser beams is commonly less than 1 milliradian, and can be much less for large-diameter beams". Not clear how "much less" is (and I couldn't find other sources), but let's say it's 2-3 more orders of magnitude. That would make it about 1 arc second (~5 microradians). The closest star is at a distance of about one parsec, which means that two photons starting here and having an angle of 1 arc second between them end up at a distance of 1 AU (astronomical unit ~ 150 MM km) between them. Assuming you can make an antenna the size of Jupiter at the target, for every photon that passes through that antenna you need to send 1 million photons. Every order of magnitude lower in the size of the antenna requires you to send two orders of magnitude more photons. In other words, an antenna 1% the size of Jupiter requires you to send 10 billion photons for each photon received. You can get around that by setting up a system of relays, but you will end up needing lots of relays, and that's a complex infrastructure you need to maintain. In any case, even after we master inter-stellar travel, don't think we'll have channels of communication of terabytes/second, every byte will be expensive to send across. Sending information in the quantities needed to 3d-print stuff at atomic level will most likely be prohibitively expensive; remember 1 gram of hydrogen contains 10^23 atoms, while one terabyte only contains 10^12 bytes. If a little latency is acceptable, it will always be many times cheaper to just send the stuff you need rather than the information to 3d-print it at atomic level. [1] https://en.wikipedia.org/wiki/Collimated_light https://en.wikipedia.org/wiki/Collimated_light

lisper 8y ago
> Sending light across interstellar distances is not a trivial affair. I never said it was. But it's a hell of a lot easier (to say nothing of faster and more efficient) than sending matter.
- credit_guy 8y ago
  Not necessarily. Remember, photons have mass. Sending a lot of them means you perform mass transfer. Since photons are not parallel, that transfer is not efficient. It's probably much more efficient to send a small package equipped with a large solar sail and push it from behind with a big laser until it reaches a relativistic speed, then decelerate it on the other side with another big laser.
  - lisper 8y ago
    Hm, that's a good point. Doing the math on that would make an interesting exercise.
    - credit_guy 8y ago
      Ok, here's the math. Solution 1: send information with a highly collimated laser. Assume 1 arc second divergence of the light beam and a distance to the nearest star of 1 parsec (the actual distance is 1.3). The diameter of the circle of light at the target is 1 AU. Assume there is a listening antenna at the target of the same size as the largest radio telescope today, which is 500m [1]. The ratio of the area of the telescope and the area of the light disc coming from the source laser is (500m/1AU)^2 = 10^(-17). You want to send about 1Tb/s of information. 1 bit/s requires about 4zW of power [2], so you need the antenna to receive about 10^12410^(-21) = 4e-9 W. You need to send about 10^17 more at the source, that means about 4e8W, or roughly 400MW. Let's make that 1GW, to account for various other inefficiencies. That's actually not so bad. Solution 2. Send a package in a solar sail powered vehicle. I don't know how to do the math, but fortunately wikipedia has some examples [3]. A randevouz with Alpha Centauri for a vehicle with a final mass of 71 tons can be done in 41 years, and requires a power of 7200 GW here and 26000 GW in Alpha Centauri (for the braking stage). Let's say out of the 71 tons vehicle at the end of the trip, only 1 ton is useful cargo, and you fill that with SSD drives. Currently the largest SSD drive is about 100 TB, and it weights about 100 grams. That's 1TB/g, or 10^6 TB/ton. That's about 100 days worth of transmission at 1 Tb/s. So it wouldn't seem remotely comparable with Solution 1. However, by the time we reach the stars, I can easily see the information density going up by a factor of 10^20 or so. So that's that. If you want to send information at limited bandwidth, you can send it using a big laser. If you can afford some latency (40 years instead of 4y), then you can send it by putting a lot of hard drives in a package. [1] https://en.wikipedia.org/wiki/Five_hundred_meter_Aperture_Spherical_Telescope https://en.wikipedia.org/wiki/Five_hundred_meter_Aperture_Sp... [2] https://en.wikipedia.org/wiki/DBm https://en.wikipedia.org/wiki/DBm [3] https://en.wikipedia.org/wiki/Solar_sail#Swing-by_maneuvers https://en.wikipedia.org/wiki/Solar_sail#Swing-by_maneuvers
  - kwhitefoot 8y ago
    > Remember, photons have mass. Really? https://physics.stackexchange.com/questions/2229/if-photons-have-no-mass-how-can-they-have-momentum https://physics.stackexchange.com/questions/2229/if-photons-...

lisper 8y ago

> Sending light across interstellar distances is not a trivial affair. I never said it was. But it's a hell of a lot easier (to say nothing of faster and more efficient) than sending matter.

credit_guy 8y ago
Not necessarily. Remember, photons have mass. Sending a lot of them means you perform mass transfer. Since photons are not parallel, that transfer is not efficient. It's probably much more efficient to send a small package equipped with a large solar sail and push it from behind with a big laser until it reaches a relativistic speed, then decelerate it on the other side with another big laser.
- lisper 8y ago
  Hm, that's a good point. Doing the math on that would make an interesting exercise.
  - credit_guy 8y ago
    Ok, here's the math. Solution 1: send information with a highly collimated laser. Assume 1 arc second divergence of the light beam and a distance to the nearest star of 1 parsec (the actual distance is 1.3). The diameter of the circle of light at the target is 1 AU. Assume there is a listening antenna at the target of the same size as the largest radio telescope today, which is 500m [1]. The ratio of the area of the telescope and the area of the light disc coming from the source laser is (500m/1AU)^2 = 10^(-17). You want to send about 1Tb/s of information. 1 bit/s requires about 4zW of power [2], so you need the antenna to receive about 10^12410^(-21) = 4e-9 W. You need to send about 10^17 more at the source, that means about 4e8W, or roughly 400MW. Let's make that 1GW, to account for various other inefficiencies. That's actually not so bad. Solution 2. Send a package in a solar sail powered vehicle. I don't know how to do the math, but fortunately wikipedia has some examples [3]. A randevouz with Alpha Centauri for a vehicle with a final mass of 71 tons can be done in 41 years, and requires a power of 7200 GW here and 26000 GW in Alpha Centauri (for the braking stage). Let's say out of the 71 tons vehicle at the end of the trip, only 1 ton is useful cargo, and you fill that with SSD drives. Currently the largest SSD drive is about 100 TB, and it weights about 100 grams. That's 1TB/g, or 10^6 TB/ton. That's about 100 days worth of transmission at 1 Tb/s. So it wouldn't seem remotely comparable with Solution 1. However, by the time we reach the stars, I can easily see the information density going up by a factor of 10^20 or so. So that's that. If you want to send information at limited bandwidth, you can send it using a big laser. If you can afford some latency (40 years instead of 4y), then you can send it by putting a lot of hard drives in a package. [1] https://en.wikipedia.org/wiki/Five_hundred_meter_Aperture_Spherical_Telescope https://en.wikipedia.org/wiki/Five_hundred_meter_Aperture_Sp... [2] https://en.wikipedia.org/wiki/DBm https://en.wikipedia.org/wiki/DBm [3] https://en.wikipedia.org/wiki/Solar_sail#Swing-by_maneuvers https://en.wikipedia.org/wiki/Solar_sail#Swing-by_maneuvers
- kwhitefoot 8y ago
  > Remember, photons have mass. Really? https://physics.stackexchange.com/questions/2229/if-photons-have-no-mass-how-can-they-have-momentum https://physics.stackexchange.com/questions/2229/if-photons-...