Well, did some testing with seastate 3(or 2. forgot), month june, surface duct, or convergence zone settings. Middle of the atlantic. Seawolf vs cargo ships starting 10miles apart.

Looks like the layer is just an adjustable "hard cap" on the sensor detection range. Kind of like detecting a cavitating ship back in SC(SCX) with the hull array. It would just pop up when it entered the threshold range.

What got my attention was playing the FFG. Noticed that dropping the towed array below a layer had no affect on ownship's SNR on the broadband. So i wondered if range+layer would degrade SNR or just cut it off. It seems to be a cutoff.

What about active sonar? Any testing done on that? Sure would be helpful to us bubbleheads when going up against a P3.

I would think that actually a Captain will do exactly that. Tactics are dominated by Physics as much as by Psychology, if not more. If by going at 130m, he has a reasonable assurance that the enemy won't detect him, that's exactly the depth he'd be at.

The enemy may also KNOW that any sub coming in at him will be coming in at 130m, but if he can't arrange for that layer to be scanned, the sub still wins. In fact, a plan that relies on a Physical Factor like this may well be more reliable than one that's based on Psychology - second guessing the enemy and hoping he won't think of what I'm thinking and come up with counters.

By having a superior intuition of the physics involved in a situation, a commander gains the advantage.

The modelling of the sonar in 1.03 is fairly complicated. The behavior can be seen in a many different ways.

What you get in one case may not at all be like in another case.

Generally speaking it is more like what you read in the publically available sonar literature with two features that have been simplified so that it can run on a PC without dragging down the processor (the sound modelling being done in the acoustics engine takes up a significant amount of processor resources as it is modelled now): it is a single path model, meaning that all frequencies from the same contact travel along a single path rather than their path being frequency dependent; the shadow zones are rather stryking so sometimes contacts will completely cut out rather than fade away. I'm not sure whether this is realistic but it seems there would be some beam spreading in real life that is not modelled.

All of these features vary greatly by environment along all of the parameters that can be set in the mission editor.

The figure Hutch mentioned:
http://www.fas.org/man/dod-101/navy/...P/IMG00015.GIF

...does make me wonder about one thing - generally speaking you could say that it's "how far between the lines"* that says how much of the sound reaches that point. In the illustration, looking at two places, right above the sub, and the place where the drawn line 'touches' the layer, it seems that right above the layer has far more distance between the lines (which should correspond to weaker sound)

*(This example should of course be done with far more lines, and make sure the 'start angle' of each was the previous +n, )

What I'm trying to say, is that based on what I've read about it, the sound should get weaker until at the layer it's completely faded, instead of being full power until the layer instantly cuts it off.

I have, however, not really that much of a clue on how the reality of it is. But if I'm wrong, then I'd like to know how my theory is off.

I think you're right MaHuJa, it's just that modelling that in DW probably would have taken up too much of the system resources to do properly. According to the developers, there is a lot going on in the background for the sound propagation modeling as it stands now.