5 or 10 meter. Makes a difference?

[jcope]

I would like to be able to use the external cam on a few tests to get a feel for how visible I really am in various conditions. I wish there was a way to go to a distant position, pull out binoculars and do try to see the sub. I often wonder whether I'm not spotted because of crew inexperience on the enemy's part or because the sub is really that hard to see.

I will definitely practice the decks awash technique. But it would be good to have a feel for how close to be when it becomes necessary, and when you're too close to even be that exposed.

I have yet to make a surface attack, except an early war shot at some docked ships.

[Sailor Steve]

Originally Posted by jcope

I would like to be able to use the external cam on a few tests to get a feel for how visible I really am in various conditions. I wish there was a way to go to a distant position, pull out binoculars and do try to see the sub. I often wonder whether I'm not spotted because of crew inexperience on the enemy's part or because the sub is really that hard to see.

The problem with that idea is that it is after all only a computer simulation. What you could see is what the computer renders on the screen. What the enemy can see is what the computer tells them they can see. This has been show countless times by players complaining that their watch crew spots a ship but they can't see it, or by the opposite - they can see a ship plainly but complain that the watch crew must be blind. Likewise the sonar - you can hear things the soundman can't, because of the way the program translates things to the screen or the speakers.

The only way going to the enemy ship and looking through his binoculars could work is if the program is adjusted perfectly so that what you see and what the AI sees is the same. As far as I know they haven't been identical yet, in any of the games.

[CapnScurvy]

Originally Posted by jcope

I would like to be able to use the external cam on a few tests to get a feel for how visible I really am in various conditions. I wish there was a way to go to a distant position, pull out binoculars and do try to see the sub. I often wonder whether I'm not spotted because of crew inexperience on the enemy's part or because the sub is really that hard to see.

A few months ago I set up a test mission with several enemy warships at different ranges around an submerged sub. Actually I had saved two identical test setups, one in daylight, the other at night. I reworked the stock "AI_Sensors.dat" file to only have one particular sensor working at a time (its tough to know what exactly detected the sub in certain conditions). My tests were to just work on the subs detection with visual AI. One during the day, the other at night, and eliminating the other possible sensors detection capabilities. I documented the amount of time it took for an enemy ship to detect the sub with a partial periscope rise; a full periscope rise; a partial sub surfaced position (decks awash); and a full surfaced position. I took into account various distances that were between the sub and target ships and attempted to see when an enemy could detect the sub, and when it couldn't. The seas were calm to remove any "Wave Factor" multiplier.

I found some surprising things.

Specific to the visual detection, I found the "Light Factor" actually is backwards to what it should be. In other words, the higher the "Light Factor" figure, the easier it is for the AI to detect the sub at night. Dropping the figure helped in preventing night detection, it also lowered the ability to have the same sub detected during the day! A sub sitting off from a warship at a very reasonable distance was not detected at all during the day. Actually, like in the stock game, they could be made blind as bats during the day, yet detect you quite well at night. Just the exact opposite of what should occur. Because there seems to be a "default" setting for the night time detection, there was no way to get away from this backward detection. I'm not surprised about this with the game, I've seen many issues that have been revealed to be wrong with the games mechanics.

Another thing is the fact that no two detections are the same. There seems to be a random generated way the sub is visually detected, even though the sub is put in front of the enemy equally. One time the sub is detected within a few seconds of exposure, the next time it may take a couple of minutes. If it's detected at all! This made my testing needing to take an "average" of the subs exposure/length of time calculation. No two are alike in detection time.

So what's it all mean?

Well, it means if you expect to have the game allow you to sneak up onto a target under the cover of darkness, you can forget it. Not without throwing off the daytime detection to the point of absurdity. It also means you need to take several tests of the same ship positions in order to get an average of detection specifics. No two are alike. This also means a players ability to slip by a specific enemy is totally random; trying the same mission another time, will probably give you a different result.

[mobucks]

I find TMO 2.2 to be reasonable in terms of enemy AI detection of my ship. Yesterday I was running at flank speed towards a small TF of 2 small cruisers. They were clearly visible with binoculars. I was running perpendicular to them at full speed undetected, but as soon as I turned and showed them my broadside, they started firing.
Night surface attacks have been made out to 1200m with decks fully awash, the AI merchants never see me on a dark night. I tried a night attack fully surfaced, I think he saw me at 1500-2000yards. I only did this once, so I'm not sure If it was detection or a random zig, I wasnt fired upon because the merchant was unarmed.

[TorpX]

Poking around in the SH III section, I noticed they have the same issues. Sometimes they have situations where their crews can see enemy ships at long distances, eventhough the enemy was blind and could not spot them. Other times the detection ranges varied drastically, due to a absurdly large random influence. They seem to be making some progress toward fixing it though.