Thanks ML for having a good look - there's nothing there I dont agree with. :up:
Perhaps we just disagree on a degree of interpretation of the results.

You dont refer to the Stock scenario Test results. I happen to think that your test results of 37 v 31.5 nm confirm my finding that the relative differential has been narrowed from Stock where one can obtain 47.9 v 37 or even (with difficulty 65.7 v 37) in that scenario.

PS. Good point about the fading I referred to that in my test reports - but that feature is something which of course is exploitable particularly with gamma adjustment.

OK, I'll go test the stock ones for a comparisson, but I'll retest on mod first since I got the SW a little late.

Ok, the good news is that we've got some progress on this discussion.

The bad news is that I'm really sick, and won't be able to work on 3.08 today, and maybe not tomorrow.

When I feel better, I'll finish it up.

Cheers,
David

Cold? Flu? Smallpox? ASWnut has your cure!:up:




*Side-effects are given "as is," and any negative effects from ASWnut's cures are in no way his fault. Use at your own risk*

Okay, I've done two more rounds of testing instead of one. Why two? Because as it turns out, the sonar model is more complex than I give it credit for, and it's effect needs to be shown for the results to make sense.

The surprise is that the louder a contact is, the less of a difference there is in sonar performance. So, when we go from stock to lw/ami, there is a variable changing that our tests haven't accounted for: the NL of the target vessel. I can't tell exactly how great this effect is, because the margin of error caused by differences in the SSP between each startup is too great.

But, I can give you some raw data:
First, the continuation of the test above:
Target contact: Victor III 5knots
TB-29 mod: 37nm
Pel mod: 31.5
Modded gap: 17%

TB-29 stock: 43.4nm
Pel stock: 35nm
Stock gap: 24%

Next, we'll try a quieter contact...like one you might actually fight against
Target contact: 688I 5 knots
TB-29 mod: 14nm
Pel mod: 6nm
Modded gap: 133%

TB-29 stock: 22.8nm
Pel stock: 16.8nm
Stock gap: 35%

I tried to run some tests to establish that the greater the NL, the smaller the difference in performance, but the data came out on both sides. Essentially, large NL differences show the effect while small NL differences show the opposite. I think that I'm right about the NL effect (the huge difference between the Victor and 688I is too significant to overlook), but that there is something else at work too. Maybe there's just a large margin or error (at least 5nm around NL=70) from the differences in the generated SSPs. Or, maybe the charts we've been relying on for NL are off a bit?

So, I can't say that in all cases, there has not been a decrease in the performance gap between the TB-29 and Pelamida, but I can say for sure that for performance against quiet modern subs, the gap has increased significantly.

I hope this puts the underlying balance concern to rest, since in game we'll be dealing with subs with NLs closer to the 688I than the Victor III.

ML I think that you have fingered an important feature the effects of which are more
relevant in SP than in MP.

The test speed of own platform is crucial as at 4kts the very faint tonal or more often an intermittent BB SNR of 1 may account for increased range results in the SW. Also sprint, drifts may be a test variable. Perhaps these may account for my Stock test ranges beeing greater and therefore showing a higher performance edge of the TB29 over Pel. Your tests confirm the 24% gap on average but higher performance is frequently obtained with a 60% gap with care. The SWs BB SNR has an advantage over the AKs
and the AKs early-day fleeting tonal ghost is no compensation.

But for MP I will have a look (this weekend) at the implications for quiet sub v sub contests. If only to reassure myself in confirming your findings. Thanks again for your testing and valuable contribution.

It only seemed fair to take time out to bring forward tests.

The AK v SW sonar game has been a major hook for many bubbleheads since SC days. So whilst ML has chosen to run the 688i against the SW I thought it may add a bit to test the AK v SW TA NB performance.

Here are my results from tests today:

Tests TA NB Range maxima.

Stock:
Ak tracking SW and 688i

688i at 18.6 nm
SW at 13.7nm

SW tracking AK
AK at 19.9 nm

Conclusion - SW has 45% increase in range over AK head to head.

Mod:
Ak tracking SW and 688i
688i at 9-10 nm
SW at 8.7 nm ( Greater ranges found in other scenarios but the differential is always steady.)

SW tracking AK
AK at 9.3 nm ( Greater ranges found in other scenarios but the differential is always steady.)

Conclusion - SW has a reduced advantage in range over AK in head to head.
The absolutes have indeed been reduced from Stock, but the differential in TA sonar performance has been dramaticaly reduced. (Test scenarios available.)

NB. Used ranged target subs and MLs run-in range reduction to obtain results. All subs at -5 knts above layer and SSP and location as before. No BB SNR readings were taken just NB tonals.(This could improve SW ranges but not material to NB comparison)

It's a good thing this discussion is entirely academic... the database can't be set any more correctly than it already is. :-?

Except for those "secret values" only Ludger and I know about. :cool: :rotfl:

Cheers,
David

PS And Bellman, your last findings directly contradict Molon's findings, so you might want to look at your scenario again, but it's really none of my business at this point... because like I said, the database isn't changing from something that's correct, to something that is incorrect... especially because this is the way things have been set for over a year and exactly zero people before this have made any comments regarding this, and believe it or not, I tend to hear about problems reasonably quickly...

Just to give MY last comment on this.

I haven't changed anything related to sonars or platforms noise levels in over a year (December of 2005).

So, I think you missed the boat, since I'm certainly not changing anything now. The SW has a VAST detection advantage over other submarines, and in some situations the mod makes this greater and in some situations the mod makes this less. These are simply the facts and the database itself is the ultimate piece of evidence.

Cheers,
David

Well, I've been corrected by LW that ownship speed doesn't matter as long as you're not in washout range, so I retract my statement about needing to retest controlling for speed. All testing was done at tactical speed and washout was not a factor.

I also want to express my agreement with LW concerning the fact that the NL's and sonar sensitity has not been changed for ages. I've been playing with these levels for a long time now, and it's always been absolutely clear that the American subs had a substantial detection range advantage due to the combination of quieter subs and superior sonar. In fact, the principal fair matchup in LW/Ami was the 688I vs the Akula II, instead of the SW vs. Akula II, because the SW just pwned the Akulas. (1.04 probably tips the matchup back to the SW).

I can only guess this issue is surfacing now instead of when the sonar sensititivies were set because of the changes in 1.04. But, what these test results make abundantly clear is that even in sonar conditions that favor long-range detection (being inside a surface duct), the Seawolf can close with an Akula-II to within no-escape range without ever being detected. And that's without playing any layer tricks. It's only a matter of time before SW skippers learn to control their speed at the right times to become proficient at doing this, and once that happens, it will be the Akula skippers bitching and moaning about the balance changes in LW/Ami--because they are going to die without ever having detected the launching platform, and CMs aren't going to defeat wireguided torps guided in from less than 8nm.

OK guys I'm going to quit the field now (general sighs of relief :D)

I am not yet convinced that something, possibly 1.04 as ML says, has'nt changed the mix.
But I'll mothball-it and perhaps return to the matter further down the road.

Thanks at least for addressing my concerns.

There were no passive sonar changes in the DW engine in 1.04.

Cheers,
David

I said 1.04 tilts the fair matchup; I didn't say that it changed anything with the sonar.

:D So who tweaked the Kilo SA ? :lol:

*yawn*

I don't have the energy to explain the difference between the acoustic engine and the interface.

Cheers,
David