Also, detected emissions (active sensors and fire control, broadcast or tight beam comms (the former far more than the latter), thermal emissions, light emissions, reflected light from other system bodies you might be near or transiting, etc) are detectable.

As a synthetic multi-sensor array, an EMS might well note a small thermal difference, note a small dark spot if the bogey crosses a light area, and a faint EM trace or some cooling signs of a drive by product all together might give the threat processor a good reason to alert on a particular area of space.

Also, flat detection ranges are a gamer thing, vs. anything vaguely real. What determines thermal detection? Quality & size of the sensor looking for it (and how fast it can process data including the original data analysis and differential analysis between data frames), the cosmic background temperature where you are looking [2], how big the differential is, how effective any limited means of temporary thermal deception are, whether the bogey has to manouver and shed heat or heated reaction mass along with that, did they launch anything that emitted, etc.

A big emission (a far away star) is a high magnitude source. So it's more likely you'll see it than a ship or even orbital station at the same distance.

Active sensors provide some to all energy that they get reflected back to their receivers. This gives them one additional advantage: They get Doppler shift data (having emitted what's bouncing back) and that gives them some speed/vector information from a bogey. But range is a big issue - energy density falls off fast with range (think of the omnidirectional radio signal as a sphere wrapped around your planet which travels outward, the sphere growing as it goes) [3]. The fall off I believe is not based on the increase in volume, simply on the increase in surface area which will intersect any bogey to be detected [4]. That means that for every doubling of distance from the emitter a bogey has, the amount of energy that hits will be 1/4 of what it was before the doubling [5]. You can see this drastically limits the ranges of active sensor use, so most of what will give you long distance info is passives. Especially since just getting the energy out there to bounce off the bogey was suffering that doubling -> quartering of power and the same would apply on the trip back. And that's assuming that the signal wasn't absorbed by ablative stuff or reflected off at sharp angles which might leave VERY little return.

Then there's the whole 'light speed' thing. You can detect planets tens of light years away (maybe longer, I haven't checked the furthest out exoplanet's distance from us) and we can see stars a LOOOONNNNGGG way off. That said, we are seeing light from them from a few years to back near the big bang, so there is no particular up-to-the-minute reports that you can get with FTL-limited radiations (doubly so if you have to create the reflection by first emitting like an active radar does, because your signal then has to travel in both directions).

I think detecting some info about a planet in another system is quite likely given late TL-9. I think you can maybe get some atmo makeup from gases, some size estimates, and maybe some density ideas if you can track orbits and the impacts of other bodies in the system. I think detecting ships near Earth likely a TL-9 late to TL-10 thing and further out through our entire system by TL-11. In the far system, you might have some more difficulty, but given the long range that would be and the travel time coming in, you could likely pick up anybody before they got close.

Would an omnidirectional radio signal emitted from Earth get to Alpha Centauri? There are about 55 doublings [6] between Earth and Alpha Centauri which means that the initial power emitted would be quartered every doubling. (I tried to get the value for the % of the original emitted power would reach Alpha Centauri, but the big number calculator I used had passed 0.00000000000000000000 by 35 doublings... so let's say 'not bleedin' much' left after 55 doublings).

A directed radio transmission with very limited limited beam scatter (or maybe a light beam with even less scatter) might manage to punch out to those sorts of interstellar distances (vs. an omnidirectional) but things like distress calls (probably not much use unless you go into cold sleep...) and so on would tend to be omnidirectional unless you had a darn good idea where you were transmitting to.

So mostly all you'll get is passive stuff from large sources and perhaps some observations of objects doing transits or having different levels of heat than the background at any great distances.

[1] Distances are dictated by many factors, so stating a 'distance' at which 'a something' can be detected would still be very subject to factors (sensor quality/size, processing power, other radiations that might interfere, etc)

[2] Cosmic background temp may vary somewhat if something locally hot is near your target or depending where you are in the galaxy (as in how close to the region of many large, hot, giant emitters, etc) and other factors for thermal. Looking for visual things, for instance, looking into a bright milky way sideways can be unhelpful, but a bright object up from the galactic plane could be easier to spot.

[3] I'm treating your emitter and its planet or ship as a point in space of 1m diameter. It wouldn't be, but in the long run would probably only reduce the number of net doublings of distance by a fraction of one doubling.

[4] Because you send the omnidirectional signal out in all directions and it is like an expanding soap bubble (I think), it is the surface area that matters, not the total volume in the growing bubble. I COULD BE WRONG. But if so, that just makes the fall off in power at the boey GREATER.

[5] Based off comparing what happens to spheres as I double their radius (the surface area quadruples).

[6] Solving A * 2 ^ N = B where A started at 1m, ended up at 3.8E16 meters (a messy approximation of Sol to Alpha Centauri) and N is what you need to solve for, and the formula for that ends up being N = ( ln(B) - ln(A) ) / Ln(2)

If you want your Trav to be a bit closer to the real world, this might give you some ideas of what TL9-12 may be capable of and what limits the real world applies.

Pardon mistakes or bad math. Some of the numbers were large and my method may have a mistake or three.

This sort of stuff would definitely be in play over TL9-10 at least, and I think beyond that in various aspects.

TomB