The Gaming Den

tzor wrote:So you got billions and billions of point sources to spectrally analyze and motion detect to see if they are moving or not and you have to do this all over your spherical field of vision. Good luck.

Dude, I do artificial intelligence. We can do that. We can do that so precisely that we can detect a star wobbling because there's a planet orbiting it. We can do that so precisely that we can detect matte-black asteroids the size of your car orbiting out past mars. We can do that so precisely that we could probably detect the Space Shuttle's maneuvering thrusters from Mars. And we can do it in a matter of hours.

Tzor, do those pictures have any relevance whatsoever? (No.) While it's sometimes hard to tell, you are a human being, not a computer. The actual device that does the scanning of the sky (which is not your squishy, human eyeballs) could detect a spaceship maintaining habitable temperatures between us and the sun. That is just something we can do. We have taken the IR shots with such exacting measurements that we really would have spotted a spaceshuttle sitting there if there had been one.

Bitching about how impossible it seems to do that is worthless, because however impossible it seems we went out and did it, and that proof of capability is a lot more important than your personal disbelief. You might as well try convincing us giant metal cans can't fly, based on how 'ridiculous' it is to try and lift tons of metal with really big fans. Even if it seems ridiculous, we went out and did it so discussion over.

Vebyast wrote:Dude, I do artificial intelligence. We can do that. We can do that so precisely that we can detect a star wobbling because there's a planet orbiting it. We can do that so precisely that we can detect matte-black asteroids the size of your car orbiting out past mars. We can do that so precisely that we could probably detect the Space Shuttle's maneuvering thrusters from Mars.

First of all, I think that it takes more than hours to verify star wobbling. While we can detect the space shuttles thrusters from Mars, we also are looking directly at the earth when this is happening, not trying to search the entire sky. Hell even the asteroids are in a generally known location.

Currently we can't even keep track of the BIG ones. NASA Falling Short of Asteroid Detection Goals

The agency is on track to soon be able to spot 90 percent of the potentially dangerous objects that are at least a kilometer (.6 miles) wide, a goal previously mandated by Congress.

To put this in perspective, the Battlestar Galactica is 1.4 km wide, so if the Galactica was in a stable orbit we would be able to eventually locate that ship with 90% accuracy .... key word is "eventually."

And I should point out that none of these are moving in non stable orbits at relatively (in terms of orbits) massive speeds.

...

Detecting a Galactica that has suddenly (comming out of warp) appeared with a signicant velocity roughly 1 AU (hint that's 8 light minutes) from your vessel requires a detection system that can dectect within minutes, not hours and certainly not days. At 50 psol the ship will be on you in 16 minutes and at the slower 10 psol in an hour and a half (of course the attack will have begun much sooner).

And I haven't even thought of initial counter measures. (The use of decoys deployed before the warp point so that the computer has to analyze not just one bogy but perhaps dozens or hundreds all around the detection field.)

tzor wrote:Detecting a Galactica that has suddenly (comming out of warp) appeared with a signicant velocity roughly 1 AU (hint that's 8 light minutes) from your vessel requires a detection system that can dectect within minutes, not hours and certainly not days. At 50 psol the ship will be on you in 16 minutes and at the slower 10 psol in an hour and a half (of course the attack will have begun much sooner).

Apparently you didn't read my link, nor did you thoroughly read yours.

The stats in my link are for a years-old consumer desktop computer; in other words, my cell phone. Then recall that we're talking about a society that has easy relativistic travel. Give me a set of telescopes and a modern supercomputer and I could easily spot the Galactica in eight minutes; give me a set of future telescopes and a future supercomputer and I'd spot it the second it came out of warp.

Your link, on the other hand, states that it's because of not having been given enough money to build a half-decent telescope, and in fact explicitly specifies that computation is not a factor. "Meeting that goal “may require the building of one or more additional observatories, possibly including a space-based observatory”", and then "The orbit-fitting capabilities of the Minor Planet Center are more than capable of handling the observations". The simple fact that you're a warship means you unquestionably have enough manpower and a good enough observatory.

Though in the future where space battles will actually take place, people will probably have developed something that makes them harder or impossible to scan.
Even though nobody right now knows how it will work yet.

We can take a complete picture of the sky in 4 hours, and the resulting image will be about 200 gigapixels. Once the image is taken, we have to do image comparison on this massive 200 gigapixel thing. The important thing to note here is that space is big and spaceships are small, so the first thing we're going to do is divide the image into subimages with overlap (because we don't want detection to fail when a spaceship is on the edge), and we can do this because no spaceship is going to be bigger than a subimage.

Let's make the subimages a megapixel (smaller than my monitor), because that's a reasonable problem for modern day regular household computers. Like mine. As a matter of fact, I'm pretty sure that task couldn't take longer than ten minutes on my machine (I'm not very familiar with image comparison algorithm running times, just ballparking it), and my machine is probably valued at a thousand dollars right now. But most of that is extraneous. The cost of the GPU in it, which is what we'd want... probably only 250 dollars right now.

So, we have 250,000 images (increased by 25% for the overlap) , each takes 10 minutes for one 250 dollar GPU to process. The sensor takes 4 hours to run its scan, so how much money does it take to build a computing bank of 250 dollar GPU's that can keep the scanner running continuously? 250,000 images * 10 minutes / 240 minutes, which is 10,000 GPU's and 2.5 million dollars. This is a 4-hour continuous scan of the entire sky capable of detecting anything in which a human being could be living by its heat, and any sort of thrust will make it shine it like a beacon. This is using modern day technology and is a very realistic expense (cost of observatory + small supercomputer and database) in terms of military spending.

Tzor wrote:Detecting a Galactica that has suddenly (comming out of warp) appeared with a signicant velocity roughly 1 AU

If we are discussing whether or not space stealth is possible and you offer, "but what if ships can teleport?" that's dumb. If you want stealth, and you're willing to violate the laws of physics to get it, how about infinitely effective heatsinks? You won't show up on an IR scan at all! Not to mention, if you're going to introduce FTL, how about introducing FTL detection?

Tzor wrote:And I haven't even thought of initial counter measures. (The use of decoys deployed before the warp point so that the computer has to analyze not just one bogy but perhaps dozens or hundreds all around the detection field.)

Differentiating between decoys and fakes is easy at a strategic level. A few seconds of analysis per decoy or fake will identify the real one, because sensor systems we have today can detect a. power plant characteristics, b. engine characteristics, c. thrust, and d. mass, and a bunch of other crap. If all of those things are not the same (power output, engine, thrust, mass), the decoys are obvious. And at that point, your decoy is literally almost as expensive as the ship itself. That's not a decoy anymore.

But really, this comment has to make me wonder; do you have any idea about the computation involved here? We're talking about doing an image comparison of more than 200 gigapixels, and you think that compared to that the difference between analyzing 1 hit vs 12 vs a 100 hits is meaningful?

Seriously, here's some pseudocode:
while (subimages left)
____do next subimage compare (runs 250,000 times, takes 10 minutes)
____if subimage compare returns difference
________do indepth analysis (runs X times, takes Y minutes)
(Huh. It collapses multiple whitespaces. Damn.)

What combination of X and Y make that matter? Even if Y is sixty minutes, 1000 decoys would be 60000 more minutes on top of 2,500,000 of computation time already required just to get through the image. That's a 2.4% increase in total scan time, and it cost you a thousand decoys that were sufficiently complicated to fool the computer and therefore expensive.

ishy wrote:Even though nobody right now knows how it will work yet.

Shut up.

See, the problem isn't that we can't think of way to do it; it's that if anyone found any way to do it, it would contradict everything we have experimentally demonstrated to be true in physics laboratories. As in, any step towards making a space ship undetectable begins with proving thermodynamics wrong. And that's not going to happen, because thermodynamics isn't wrong. I can demonstrate that 2!=3, and nobody's going to magically find a way to make 2=3 in the future because that is an impossible assertion and we already know that.

DSMatticus wrote:Shut up.

See, the problem isn't that we can't think of way to do it; it's that if anyone found any way to do it, it would contradict everything we have experimentally demonstrated to be true in physics laboratories. As in, any step towards making a space ship undetectable begins with proving thermodynamics wrong. And that's not going to happen, because thermodynamics isn't wrong. I can demonstrate that 2!=3, and nobody's going to magically find a way to make 2=3 in the future because that is an impossible assertion and we already know that.

Developing a weapon that will destroy/disable/jam enemy sensors is impossible? How does jamming a signal violate the laws of thermodymanics?

Winnash wrote:Developing a weapon that will destroy/disable/jam enemy sensors is impossible?

You know, that sounds totally reasonable when you pose it in such a generic way as to have no scientific meaning at all.

1) Destroy. This isn't stealth; you've targeted enemy equipment and blew it up. You just started combat. If I walk up to you and stab you in the eye, I am not being stealthy. Though, it is a very effective combat tactic, and it will make stealth in the future easier, but presumably if you can blow up all their sensors you've won anyway.

2) Disable. See above.

3) Jam. Here we can get into actual scientific discussion. So, let's talk about how the enemy is detecting you, and how you can stop that. They are detecting you because you give off heat and brightness different than the background heat and brightness of space. First, you could try raising the heat/brightness of the surrounding universe to mask your position, but this isn't really stealth, they're going to see this blob, and they're going to know it's deliberate and that someone is inside it. You've made yourself obvious, but you have masked your exact position. Except why bother? We're talking about seeing things light minutes away; if you want to mask your exact position, hook your maneuvering thrusters up to a RNG and let the lag do the rest. Nothing they shoot at you will hit. Alternatively, you can lower your heat and brightness to match the surrounding universe. Which means any organic being on your ship freezes to death, you turn off the power plant, and you stop using the engines. This isn't exactly practical.

Now, if you want real stealth where you are undetectable AND your ship is actually running, physics says no. Because your running ship is hot, and space isn't, and the second you emit any kind of heat you are detectable.

The closest you are going to get to jamming is to confuse enemy sensors with overload. Think about the way a plane will deploy a crap load of flares in the movies to confuse a heat seeking missile. Pretty much the same general strategy.

You generate many different heat signatures around you. This does not qualify as stealth, it's like setting off 15 flash bangs in various places to cover the sound of your footsteps. They might not know where you are, but they sure as hell know something is there.

The problem with this, is that as stated previously, enemy super computers are going to cut trough the decoys in seconds. That means that the only tactical application of this type of system is to cause a fuck up in enemy firing solutions for a split second.

DSMatticus wrote:Now, if you want real stealth where you are undetectable AND your ship is actually running, physics says no. Because your running ship is hot, and space isn't, and the second you emit any kind of heat you are detectable.

Generally, I agree with you, but this is wrong. You can be a chameleon just fine, but you have to store the heat in some sort of sink. Like Mass Effect, but without the benefit of unobtainium. That probably means that it's not feasible on small vessels, and probably isn't feasible 360 degrees either. It's also not a long term solution, unless you want to cook your crews.

The heat sink is still hot and detectable. Unless you periodically jettison it, or it's some sort of heat radiator that trails behind your ship on a cable, people will still be able to tell where your ship is by looking at your heat sink.

Grek wrote:The heat sink is still hot and detectable. Unless you periodically jettison it, or it's some sort of heat radiator that trails behind your ship on a cable, people will still be able to tell where your ship is by looking at your heat sink.

Put a big hemispherical shield on the front side of your ship. Tether it a ways in front of you with fairly well insulating pylons. Run a heat pump hard, all the time, and heat/cool your shield from the back so it blackbody radiates at an intensity you choose. Pump the excess heat from this (there will be a lot of it) to the rear of your ship. You can now set your signature to whatever you want, for observers who are mostly forward of you. You stand out like a damn beacon from the rear, but that may be acceptable.

Choose a signature based on what the background looks like to the rear of you. Don't do something stupid like flying in front of the sun.

You are now camouflaged! It's very directional, depending what the sky looks like behind you, but it's definitely not nothing.

The easy thing about sci-fi battles is that you can make up whatever the hell you want.

Obviously you're probably using FTL tech which is something we don't have right now anyway, along with energy shields and whatever else fun stuff you may want.

Now however you want space battles to go is up to you, you want to decide what you want them to look like first, and then go ahead and make it happen. Maybe you can have stealth ships because you're not using a hot rocket thruster but rather some kind of hyperdrive engine that partially phases the ship into another dimension, making it difficult to detect. Detecting it is actually done with complex hyperspace sensors, which in turn have arbitrary means of jamming them. Assuming you want to get around the infrared sensor. Or you could just go with Gundam style Minovsky particles that jam known methods of detection.

Doesn't really matter.

For everyone saying that stealth is impossible in space, remember by modern science, so is FTL and a bunch of other sci-fi tropes. Don't let modern science get in the way.

Heatsink is the wrong word here. Heatsinks are all about giving heat off; their entire purpose is to exchange the heat of your components with the not-heat of the medium they're in. Like air, in a computer's case. They're going to be incredibly visible.

What you're actually asking for IS unobtainium, because it is a impossible substance that simultaneously absorbs your ship's heat but does not radiate heat of its own.

Now, there's the idea of a giant shield in front of you that you pump heat from. The problem with this is that you are now running a heat pump. Which means you are turning on the power plant. Which generates more waste heat which must be pumped away. Which requires running the heat pump harder. Also, narrow emissions? Efficiency goes down as the emission cone tightens. Moral of the story? Your ship is huge and nearly all radiators. Your best bet is to not attach to the shield at all, or to have an attaching/deattaching scheme where you can hook up to adjust it as needed. But you're still running a heat pump into a narrow cone 24/7, and now you have to seriously worry about obscuring enough background light that you're visibly detected. And it's still going to pick up your heat, just more slowly, so you are adjusting it.

But this is all irrelevant, because this sort of camouflage has a very narrow effect range and it's incredibly unrealistic to expect the enemy to not have a huge range of vision. Really, your camouflaged ship is huge thanks to those radiators, and space journeys already take forever, so you're not getting anywhere fast. Detecting you is a non-urgent task, and that makes the solution cheap. You fire off sensor probes in every direction, outward from the area you want to observe. The probes are low-performance, but they're only interested in looking inward at the area where things are actually happen, and their mission isn't time critical, so who cares? They don't have supercomputers attached, instead they just beam the information back to a central location, which processes it and beams out information to relevant field ships. It takes about a dozen probes before there is nowhere you can realistically vent your emissions to not be detected, no matter how big your 'invisibility shield' is.

Swordslinger wrote:For everyone saying that stealth is impossible in space, remember by modern science, so is FTL and a bunch of other sci-fi tropes. Don't let modern science get in the way.

That was kind of the point. When we say impossible, we aren't talking about an inability to imagine it, we're talking about your an inability to feasibly do it. If you want technomagic, you can do a lot.

Holy crap. I did not expect this conversation to conform this closely to the Atomic Rockets prototypical space stealth discussion. I mean, seriously:

Ken Burnside (whoever he is) wrote:Most of the arguments on thermo and space detection run through a predictable course of responses:

• 1) "Space is dark. You're nuts!"
  2) "OK, there's no horizon, but the signatures can't be that bright?"
  3) "OK, the drive is that bright, but what if it's off?"
  4) "But it's not possible to scan the entire sky quickly!"
  5) "OK, so the reactors are that bright, what if you direct them somewhere else..."
  6) "What if I build a sunshade?"
  7) "OK, so if I can't avoid being detected by thermal output, I'll make decoys..."
  8) "Arrgh. You guys suck all the fun out of life! It's a GAME, dammit!"

I suspect DSMatticus of cheating and reading the script.

---

Also, DSMatticus, you've ballparked completely wrong; each image stitch takes about a thousandth of a second. It's so fast, in fact, that we can do it in realtime on point clouds. Since the process is disk-bound and embarrassingly parallel, we distribute the data and processing out over a small hadoop cluster stuffed with HDDs to parallelize disk I/O as much as possible; the entire thing could take as little as thirty seconds. This is what NASA is doing right now, and it's why their only limiting factors is the number of tubes they can point at the sky.

DSMatticus wrote: That was kind of the point. When we say impossible, we aren't talking about an inability to imagine it, we're talking about your an inability to feasibly do it. If you want technomagic, you can do a lot.

Mainly I'm just confused at what kind of setting you're talking about.

It sounds at times like you're talking modern tech, which is kind of silly because space warships as a whole are pretty unfeasable based on current technology. We are still waiting for a leap in technology to even make space travel a productive endeavor. A modern day space battleship would just be an anti-satellite weapon, because that's the only thing in "space" worth shooting at.

But once we get the leap in technology to make space travel worthwhile and profitable (asteroid mining and such), we may well also have some technological advancements to make stealth technology work. I seriously doubt heat producting thrusters are going to give us that level of space travel. They're just too inefficient. We may well come up with some kind of cold propulsion system, because the heat energy right now is just waste product.

That's not true. Space is filled with aliens who want nothing more than to kill us because they don't like our faces. The SciFi channel told me so.

Swordslinger wrote:It sounds at times like you're talking modern tech, which is kind of silly because space warships as a whole are pretty unfeasable based on current technology. We are still waiting for a leap in technology to even make space travel a productive endeavor. A modern day space battleship would just be an anti-satellite weapon, because that's the only thing in "space" worth shooting at.

But once we get the leap in technology to make space travel worthwhile and profitable (asteroid mining and such), we may well also have some technological advancements to make stealth technology work. I seriously doubt heat producting thrusters are going to give us that level of space travel. They're just too inefficient. We may well come up with some kind of cold propulsion system, because the heat energy right now is just waste product.

The bits about infrared camoflage are hard physics. There is no way for space stealth to both exist and agree with our experimental observations. There isn't even room for a special-relativity-replacing-newton's-laws equivalent, since we have observations and Theories that describe the regime we're discussing as exactly as we can measure. When he talks about "technomagic", he really means "a wizard did it".

The parts about sky scanning use modern technology to drive the point home just how easy it would be for an advanced civilization to do the same thing.

---

KaNT wrote:That's not true. Space is filled with aliens who want nothing more than to kill us because they don't like our faces. The SciFi channel told me so.

lol.

You're missing the point. Although chemical rockets obvious produce a lot of heat, so does trivial stuff like life support. And since we can notice that kind of thing now, we can really notice it in the future. The essential physics of the situation is that it's vastly easier to detect something then it is to hide it, and using a bullshit FTL method doesn't change that at any point before the one where your entire setting is bullshit. Since this thread is explicitly about calling people on that sort of bullshit, you can't get away by waving your magic pixie wand of 'future tech'.

Swordslinger wrote:once we get the leap in technology to make space travel worthwhile and profitable (asteroid mining and such)

We have the technology to do this now. We just don't have anyone willing to invest in such a (seemingly) risky business proposal. Intrastellar space travel isn't difficult, it's just super expensive. Only travel between stars is at all difficult from a technological standpoint using modern technology.

we may well also have some technological advancements to make stealth technology work. I seriously doubt heat producting thrusters are going to give us that level of space travel. They're just too inefficient. We may well come up with some kind of cold propulsion system, because the heat energy right now is just waste product.

This is just absurd from a physics standpoint. The only way you get places in space, sort of crazy unrealistic technobabble about reactionless drives is to shoot shit out of the end of your rocket so you get pushed the other way. That fact isn't changing any time soon, and probably will be true forever.

We are talking about realistic tech, as in things that are potentially way off or maybe possible right now, and have tested scientific principles backing them. The sensor part is right now, with modern technology.

Swordslinger wrote:we may well also have some technological advancements to make stealth technology work.

And this is the part that's being argued, because it turns out that's impossible unless the technological advancement in question is, "fuck you, physics." You are a human being, and as such require a certain amount of heat to live. Similarly, there is absolutely no device ever that will be 100% efficient and not produce heat as a byproduct, and by that token even an unmanned spaceship produces heat. Heat radiates, and is passively detectable. And not surprisingly, it turns out space is really, really super cold, so pretty much any heat at all in space shows up really, really easily. And therefore pretty much everything provides us a nice little heat image that makes it trackable. This includes asteroids, which don't even have power plants, engines, or power, and if we can spot the heat they radiate what possible hope does a spaceship have?

Now, what you can do is limit your heat emission profile. Fectin is talking about that, and it can get close. One way of doing this is to get a radiator like you see on your CPU (probably) - all those parallel planes. And then you stick that directly towards the enemy such that the radiated heat is so narrow as to be undetectable. Another is to try and fire your emission in a narrow cone behind you, such that your own ship blocks it. Another is to put an actively cooled plate in front of you to mask your radiation. Some of these synergize, but the main point is that they are only camouflaged from a particular side, and they require you know exactly where your enemy is and that he is only one direction from you with very low tolerances for deviation.

Grek wrote:We just don't have anyone willing to invest in such a (seemingly) risky business proposal.

It's also not profitable yet. Anything that could be found on an asteroid is currently available on Earth cheaper than cost of mining it from the asteroid field, even ignoring the initial investment costs. Which means as a business plan, mining the asteroid field is on par with "we're going to burn the investor's money in giant hobo-style drums."

One thought does occur to me with regards to the heat issue. the thermodynamic problem is that all of the necessary systems generate heat, but that heat is essentially a byproduct, it's waste heat. so if we discount any unobtainium based way to hide the heat of your ship, your detectability is directly proportional to energy efficiency. of course, you're already matching the background heat even if you have perfect (thermodynamicly impossible) efficiency, so the real question is how big a variance from the background is required for detection? is 90% efficiency enough? 99%? I'm seriously asking, It's been years since my last physics class.

Zinegata wrote:Read Atomic Rockets. Real space combat will be terribly, terribly boring and will largely involve using a rocket's thrusters as a weapon against enemy ships.

I fail to see how that's boring.

Anyways, I showed my cohort the thread and we've come to a couple conclusions I think will make the setting more internally consistent while allowing for a variety of possible stories and scenarios, which is all I was really after. Thanks for the help and resources.

I'll go back to never posting, ever.

DSMatticus wrote: It's also not profitable yet. Anything that could be found on an asteroid is currently available on Earth cheaper than cost of mining it from the asteroid field, even ignoring the initial investment costs. Which means as a business plan, mining the asteroid field is on par with "we're going to burn the investor's money in giant hobo-style drums."

The money in asteroid mining is not in shipping stuff back down to earth (which is ecologically damaging and potentially an existential risk if some doofus miscalculates how much damage slamming all that metal into the curst will do) but in taking the metal and cold welding it together into satellites, solar panel frames and other things that people would pay money to own once they're out in space but are prohibitively expensive to build on earth and then launch into space. Your big seller isn't iron or nickel or some shit, it's rights to your solar energy/communications satellites.

norms29, that discussion begins with unmanned space vehicles, because that heat waste byproduct is absolutely essential for organics like us to not die some kind of death, most likely painful. Otherwise, someone linked to the exact equation you need for this earlier in the thread. They helpfully mention some drive efficiencies, but I feel like I need to call BS on them. Can we really get chemical rockets up to an efficiency of .95? But yeah, if their numbers are right (and they probably are, despite my disbelief), then no, 90% is definitely not enough, and 99% probably isn't either. As a matter of fact, you're probably not going to find anything efficient enough, ever.