Question on sensors

Sensors rely heavily upon brightness... and the engine's brightness caculator always seemed a little wierd. Whenever I mucked around with it I got unpredictable results.

It occurs to me that signal dampers were never accurate either. (At one point I had problems with ships being overbright, and attempted to use signal dampers to modify.)

I still believe we don't have the total formula for brightness. Some multiplier exists... distance, cumulative additions of system brightness, or something else.

<font size="1"><font face="Book Antiqua"><font color="black">"Never before in the history of the world had such a mass of human beings moved and suffered together. This was no disciplined march; it was a stampede-- without order and without a goal, six million people unarmed and unprovisioned, driving headlong. It was the beginning of the rout of Civilisation... of the massacre of Mankind."
--H. G. Wells The War Of The Worlds</font id="black"></font id="Book Antiqua"> </font id="size1">

I do know that ship speed increases brightness. A ship stopped will be at min brightness, and a ship at max speed will be at max brightness. This was a bit counterintuitive for me. But AI ships don't have reactors - so they couldn't use heat generated to create brightness. Which means you can't do a Firefly like manuver of coasting in with reactor powered down.

Your right - all kinds of things could factor in. CPU, occlusion monitors, reactor, damage levels. We do know it can work the proper way, with stations, under some circumstance.



-Gtrout

There are all kinds of brightness modifiers in the game and they all effect the "sensed_brightness" of a given ship. The value is very dynamic.
(Just a thought - engine/power induced brightness could be a logrithmic value and not linear? Maybe sensor values are logrithmic in general???)

I would have guessed AI doesn't fire anyhow, unless they can at least identify the faction of the other ship? Why bother with the ships name/ID?




Iwar2 Multiplayer Fan Site

I should run a test of linear vs log. Good idea. POG does have support for both, so we know PS thought both were important. It is hard to know what they ended up using. The tests so far follow this formula SensedDistance=(target_brightness - sensed_brightness) * sensor_range.

The reason for wanting the better model is writing traffic scripts (or guard combat scripts). I would like to know when the patrols orbiting the station can clearly identify a ship, and then might choose to attack.

A stealth dual can happen where both sides are moving to get a clear id on the other side first. Or even to keep some ships back out of sensor range and launch missiles "over the horizon". It would work better if ships were seen, but could not be identified. It gives the player a target blip to watch and creep up on. Otherwise ships just jump out of the darkness. Which can work, just not as nicely.

And I would like to write some ship control code that does not allow AI ships to track a player target half way across the star system. Currently it appears that once the AI is given an attack order, it never loses sight of the target. Which is not true for the player. And makes escaping too hard.

-Gtrout

Originally posted by GrandpaTrout

Currently it appears that once the AI is given an attack order, it never loses sight of the target. Which is not true for the player.
-Gtrout

Yes, that seems to be true and is annoying. Perhaps AI always knows the players position because AI is much "closer" to Flux?
PS may not have implimented a chase interuption for reasons of simplicity. The generic traffic scripts are based on the players location and not on guard/patrol orders around a specific location. There is probably no handler that stops them or sends them back to thier patrol area (cull them and wait until the player reapears at the location in question).
Since you're making the effort and implimenting a handler anyhow, you could make your own rules: Maybe just use a more or less static distance range and factor in the the player ships max_brightness and min_brightness before calling off the "hounds". (where distance could be either between Base and AI or player and AI, or both)

I haven't messed with sensor range much. In the MP game it is simplified and several of the modifiers are not active. There, the basic detection ranges do appear to be logrithmic in relation to the max (or was that min brightness?) values given in the objects ini file.





Iwar2 Multiplayer Fan Site

I decided to take a shot at the sensor problem after finishing my little theory on “the best target for the situation� for a ship in combat. (I can post it if you want.)

I started by using the convenient ship and station at Bedalov, courtesy of the MyEnv package. I had passive sensors on my ship with a range of 300k and an identification range of 100k

I LDS’d over and looked for when ID came up, ship came up, etc. For the dead cruiser (modified to brightness and min brightness of 1), it popped onto the sensors at around 270k. I realized that I forgot to set the sensed brightness on my sensors to zero. I flew out to around 400k and targeted the nearby planet to clear off the station from my list. I turned around and manually LDS’d in. The station popped in as unknown at 300k and was ID’d at 100k. So apparently sensed brightness does not affect stations.

I set sensed brightness to 0, and decided to see what happens when range is less than identification range. I set range to 100k and identification range to 300k. This time, everything showed up at 100k, all ID’d.

So from what I can tell, ships always show up at range and identification range only applies to stations. Sensed brightness has no effect on stations.

To the AI, what’s the difference between an unidentified ship and no ship at all? Neither gets shot, and that’s all the player will see. Just set the AI’s sensor range to what you want its ID range to be, use Sim.DistanceBetween to find if a ship is in the unidentified area, and write a script to have the AI fly over for an ID.