SUBSIM Radio Room Forums - LWAMI Playtest One Now Available at the CADC! ---UPDATED---

SUBSIM Radio Room Forums (https://www.subsim.com/radioroom/index.php)

- Dangerous Waters (https://www.subsim.com/radioroom/forumdisplay.php?f=181)

- - LWAMI Playtest One Now Available at the CADC! ---UPDATED--- (https://www.subsim.com/radioroom/showthread.php?t=93264)

Quote:

- please change UUV passive sensor form VLF to HF passive sonar. We don't need to see low frequency lines from it, it has no display, it's an AI sensor in fact, so we CAN make it realistic in terms of frequency coverage and det ranges

Game engine ignores such values for the UUV. I left it as it was for reference.

In terms of the sonar performance... I'm quite surprised by this. I tuned it so that it would never detect a contact before the sphere arrays...

In terms of the sensor coverage, I left it the way it was, I assumed this would be discussed.

The sensor performance you describe is a big surprise and is unlike anything I've seen.

Quote:

Game engine ignores such values for the UUV. I left it as it was for reference.

I'm surprised. It's possible to change UUV sensor sensivity, but impossible to change frequency range ? Have to try it myself... right now, I'll set it to 1100-1200Hz and see if it reduces det ranges.

The mission I run, was simply Capitallists vs communists - there is Rubis at 7kts (non cav) which was detected by UUV right out of the tube from 20nm...

edit: I changed UUV sensor freq range to 1100-1200Hz and in the same mission it didn't detect anything (but me), and there were two cavitating Akulas and Type-206 10nm from me. Now reverting to 0-1200Hz...

Cavitating Type-206 detected from 28nm, Rubis 7kts from 15nm, many other contacts, just after leaving tube.

Yeah I know should test same scenario, but from above seems that what is set in frequency limits is making difference in UUV performance.

Now changing to 1200-12000 range...

OK, I have setup with some close targets, Rubis 7kts at 10nm. Launching UUV... didn't detect it. It even didn't detect a close freighter from 6nm !!!

LW, db freq range definitely works for UUV sensor for me. At least this is what I can conclude from above. Two times nothing detected with high freq limit (1100-1200 and 1200-12000), lots of targets, even far away, detected with stock 0-1200 freq limit.

P.S. Cross layer trawler not detected at 5nm, detected at 3.6nm, 1200-12000Hz freq limit. At what range is 8kts trawler detected by 0-2000Hz UUV sensor ? Freighters were even at 10-15nm if not longer...

P.S. Now, when I believe that frequency range is in fact working, I'll try to use DWAnalyzer to give some results for comparison. As long as there is no layers involved, DWAnalyzer sonar model is still very good.

Sea state 1, convergence (but above layer, target and sensor both at 60m/200ft, rock bottom, DWAnalyzer results, all tests on LwAmi302 database

Playtest One UUV, +7 sensivity, 0-1200Hz freq range

Akula II 10kts 180 meters
Akula II 5kts 12m
Kilo Imp 5kts 9m
Kilo Imp 0kts 5m
Rubis 7kts 300m (very strange, maybe mine was CZ contact ?? anyway it is detected by UUV from 10-15nm in every my test in CvC scenario)
Han 5kts 6500m
Han 10kts 12130m
Trawler 5kts 6700m
Cargo Ship 10kts 24000m
Cargo Ship cav 33000m
Supertanker 15kts 29000m
Car Carrier cav (most noisy sound in db) 43000m

so loud targets are detected from far ranges

Now +7, 1200-12000Hz sensor:

Akula II 10kts 10m
Akula II 5kts 4.8m
Kilo Imp 5kts 3.7m
Kilo Imp 0kts 2m
Rubis 7kts 213m
Han 5kts 1922m
Han 10kts 4630m
Trawler 5kts 2333m
Cargo Ship 10kts 10133m
Cargo Ship cav 16300m
Supertanker 15kts 12133m
Car Carrier cav (most noisy sound in db) 21850m

Now +7, 3000-12000Hz sensor:

Akula II 10kts 9.6m
Kilo Imp 5kts 3.9m
Rubis 7kts 212m
Han 5kts 1922m
Han 10kts 4024m
Trawler 5kts 2012m
Supertanker 15kts 12074m
Car Carrier cavitating 19230m

Now for comparison -3 (!!!), 3000-12000Hz sensor:

Akula II 10kts 250m
Kilo Imp 5kts 200m
Rubis 7kts 1920m
Han 5kts 6700m
Han 10kts 9400m
Trawler 5kts 7300m
Supertanker 15kts 18100m
Car Carrier cavitating 25500m

results for stock db -3, 0-1200Hz UUV seeker - later

P.S.

Hm, found something really interesting. I couldn't further decrease detection ranges by increasing freq limit over 3000Hz, even for 20000Hz ranges were the same. So I checked how frequency absorbtion values (visible in DWAnalyser) are looking for different frequencies. And I'm little disappointed.... Seems there is no frequency absorbtion function, only four discrete bands with fixed freq absorbtion coefficient value for each...

Results of freq. absorbtion for passive sonar at 50000m:

Hz 50km 1m (min value)

0 38 4
20 38 4
40 38 4
49 38 4
50 38 6
59 38 6
60 55 6
80 55 6
100 55 6
300 55 6
500 55 6
800 55 6
900 55 6
909 55 6
910 71 8
1000 71 8
1200 71 8
1300 71 8
1400 71 8
1477 71 8
1478 81 8
1480 81 8
1500 81 8
2000 81 8
3000 81 8
5000 81 8
8000 81 8
12000 81 8
20000 81 8

edit: I wonder if this changes with depth or conditions... maybe another time... no, checked this - at least in DWAnalyser water depth doesn't change frequency absorbtion value. So that's why in very shallow water we can get better propagation than on deep water !!!! (narrow sound channel, yes, but I believe in RL multiple bottom bounces should weaken and disperse the sound quickly, and the very low frequency sound waves maybe can't propagate at all when the sound channel is smaller than sound wave length ??). So that would be next thing to ask Sonalysts for - after discussing it with our experts here - to make frequency absorbtion value depending also on water depth and bottom type.

As can be seen, there are only 4 different values for whole range 0-20000Hz... Freq bands would look something like that:

0 - 59Hz 1 or 1.52dB/km
60 - 909Hz 1.45 or 2.2dB/km
910 - 1477Hz 1.87 or 2.84dB/km
1478 - 20000Hz 2.13 or 3.24dB/km

little strange... comparing this to this graph:

http://www.fas.org/man/dod-101/navy/...P/IMG00005.GIF

on the graph in 1000-10000Hz range, freq absorbtion changes 100x !! At 10kHz it's one dB for 1000m, at 1kHz it's only one dB for 100.000m. Is that graph good ?

And we have... one freq. absorbtion value for this band !!! Any thoughts ??

According to the graph, between 100Hz and 1kHz real frequency absorbtion in water increases about 100 times. 100 times (in dB already - because coefficient value is in dB/m). In DW seems that between 100Hz and 1kHz there is difference of only about 1.44-1.87 times... :o What's up ?

In original DB there are values 100, 300 and 800Hz used. From data above it would seem that 100 and 800Hz are valid but 300Hz would not make any difference as it comes into same band as 800Hz... This has to be checked in gameplay tests - is there is ANY difference between 300 and 800Hz sensor...? BTW the minimum possible (at 1-100m distance) frequency absorbtion values are greater than zero (-4 at 0Hz to -8 above 1000Hz) and this value has only 3 freq bands... How there can be frequency absorbtion loss of 4 to 8 game dBs (8 to 16 real life dBs) at 1 meter distance ??? All this is little complicated, I'm far from understanding fully how this works in DW :-/.

So in fact there are only 4 sonar "performance" bands, depending on sonar min frequency value. Anything higher than 1500Hz doesn't make any difference. I would suggest to set UUV sensor min. frequency to 1500Hz to get the last, higher frequency absorbtion.

P.S. While searching for frequency absorbtion graphs on the web, i've found this:

http://www.npmoc.navy.mil/KBay/oceano.htm

Absolutely fantastic site with lot's of sound propagation examples and info !!

Huh?

I tested the sh*t out of this...

Let me say for the record that this whole UUV business if beginning to thoroughly p*ss me off.

Just one frustration to the next...

I tested the UUV at 1200-2000 and no change what so ever.

I tested the UUV from 1800-2000 and I get something difference.

Just like there is a fairly predictable sensitivity curve up to 7, and then after 7, it just drops way the heck off.

There is no good predictable way for this behavior to be calibrated with the other sensors... a player sensor with sensitivity 1800-2000 and 7 would be basically useless. I'm just going to have to do this all from the seat of my pants and test it with every kind of contact.

Every time this is run, its giving someone a different result from the next person, based on some minutae BS.

Obviously, if I thought it would do anything, I would have changed the frequencies sensivities.

So everyone just ignore the UUV sensor for now. When I'm not really frustrated by this whole thing, I'll work on it again. :nope:

Quote:

Originally Posted by LuftWolf

a player sensor with sensitivity 1800-2000 and 7 would be basically useless.

Of course when you reduce frequency band, you can simultaneously increase sensivity by few points to restore planned performance for quiet contacts. And range for loud contacts will be still shorter than before.

Well UUV sensor is in fact AI sensor (no display, detections handled by AI) so can't be compared directly to human player sensor specs...
I see dramatic difference in performance between 0-2000 and 1200-12000 or 3000-12000 sensor. To be sure, I'll try tomorrow to make a test mission (instead of taking CvC one) and get clear results (det ranges for both) and then send you the mission for you to check if you get the same results as me. My new specs for UUV passive will be 1500-12000Hz (however I think 1500-2000Hz sensor would work exactly the same...)

Such sensor would be impossible for human player platform, as there is not many discrete lines above 1500Hz, so in most cases there would be nothing on NB display to identify :-). Only BB would show trace...
But for AI seems to be no problem - even 3000Hz up sonar works without problems, detects targets even though there is no discrete lines above 2000Hz in the game, highest are around 1800Hz.

Really, I never even thought about that! :o

That's great news! :D

Thanks Amizaur.

Cheers,
David

BTW, you have the 1.03 Analyzer version? :doh: :o :up:

May I have that please? :)

Cheers,
David

Quote:

edit: I wonder if this changes with depth or conditions... maybe another time... no, checked this - at least in DWAnalyser water depth doesn't change frequency absorbtion value. So that's why in very shallow water we can get better propagation than on deep water !!!! (narrow sound channel, yes, but I believe in RL multiple bottom bounces should weaken and disperse the sound quickly, and the very low frequency sound waves maybe can't propagate at all when the sound channel is smaller than sound wave length ??). So that would be next thing to ask Sonalysts for - after discussing it with our experts here - to make frequency absorbtion value depending also on water depth and bottom type.

Amizaur, in this case, I'm not sure the analyzer is kicking back the correct result... for a shallow rock bottom limited environment, the signal travels quite far, but for a mud and sand bottom, the distances drop off very rapidly.

The rock bottom limited environment may transmit signals better than the top level of a surface duct, but the mud and sand bottoms create a very bad acoustic environment, especially in shallow water. Am I understanding your analysis correctly?

Amizaur, the frequency ranges for the UUV aren't doing anything on my computer. :-?

The good news is that it really doesn't have to matter... as long as I can set the NRD the way it needs to be, the actual detection ranges will be the same, since DW isn't using a multichannel model anyway, just using an algorithm to determine signal loss as a function of frequency... SO, if I change the gain of the sensor capable of detecting 50hz signals to detect them at a range equivalent to the sensor with a higher signal gain detecting only the same contact at 1800hz, the end result will be exactly the same, in DW terms.

Gosh this has been any annoying problem. Ok, so the end result is that I'm going to reduce the sensitivity of the UUV even further from where its at, perhaps to about +15.

I have posted the LWAMI Playtest One WITH ATP correction to the CADC:

http://www.orionwarrior.com/forum/sh...ad.php?t=29521

From the playtest readme:

Quote:

To Install: Unzip the file into your main DW directory allowing the unzip program to overwrite all files and install to the correct directories.

This playtest should be considered a beta. It combines LWAMI 3.02 with the following changes:

ADDED IN THE ATP VERSION

UUV-The UUV sensor sensitivity has been reduced and given a hardcap of about 20nm. Also, the range of the UUV is now greatly reduced by running it at high speed, especially at 20kts. Also, and I just noticed this although it appears on my computer to be in the stock game as well, the course of the UUV is going to be off whatever is indicated in the fire control panel controls by the firing angle of the torpedo tube that fired it... I can't correct this, but I tried, although it doesn't seem like a big deal to me. :-P

UGST and ADCAP- I have tuned the wirewatch sensor and doctrines. The only ownship condition now monitored by the wire is the distance from the launchpoint of the torpedo... this is measured as a radius. I can't have it be the truerun of the submarine because the doctrine isn't getting good info from the Sim. Again, this is fine in my opinion because it allows the submarine some loiter capability. Also, the Enable distance and the wiredistance for the torpedo are now measured as the true distance run rather than the radius distance from launchpoint. AND THE MOST IMPORTANT THING: the Advanced Torpedo Physics have been implimented for the ADCAP and UGST, so that means they will run farther when slower and running more shallow (although the database still needs to be updated to fully take advantage of this at the extreme long/slow range/speed settings).

END OF THINGS ADDED IN THE ATP VERSION

Advanced Torpedo Control Mod for UGST and ADCAP:

Set the torpedo to fire as normal. If you don't use the wire commands, it will behave as always. The wirecontrol commands are now as follows. Note, you must wait at least a second of game time between clicks, but it is possible to hit the button several times to go through the cycle quickly. A single preenable click will preenable the torpedo if it is enabled or do nothing. A second preenable click will send the torpedo to the preset search depth. A third click will send the torpedo to the ceiling. A fourth click will send the torpedo back to launch depth. The enable button works as follows. A single enable click will set the torpedo speed to 40kts and enable the passive sensor. A second click will enable the active seeker on the torpedo and set the speed to max speed. Further clicks of the enable button have no effect on the torpedo unless it is preenabled again, and then the behavior is reset to the beginning of the enable cycle.

WireBreak Mod:
Wires are now limited in range and ownship maneovering parameters. The ADCAP has a 10nm internal wire and a 5nm wire on the launching platform. If the torpedo or ownship travel farther than those distance FROM THE LAUNCH POINT, or if the range between ownship and the torpedo exceeds that distance, the wire will break. The UGST has a 25km internal wire and a 5km wire on the launching platform. Additionally, if the opening speed between ownship and the torpedo exceeds about 60kts for the ADCAP or 55kts for the UGST, with ownship movement accounting for no more than 20kts of that calculation, the wire will break. These maneovering measurements are unintentionally fuzzy, however, it is something that occurs naturally that I was going to build in anyway, so it works nicely. :-) What this means in practical terms is that a slow running torpedo gives the launching platform much more flexibility in maneovering the ship, whereas a torpedo running at maxspeed is much more prone to a maneovering-related wirebreak. NOTE: When the wire is broken, its broken. However, sometimes the interface will momentarily display the torpedo as preenabled, but it reenables soon enough not to effect the game in any way. The only unfinished part is that you can still shutdown the torpedo even after the wire is broken... we can't take this out. However, it is very minor, in my opinion, seeing as the user would typically reload his tube after the wires are broken anyway. ;-)

Advanced UUV Mod:

The UUV is much more quiet now, and is very hard to detect without cavitation. The passive sensor has been reduced in sensitivity considerably and the active sensor has been disabled completely (mostly because its broken in DW 1.03). The UUV now has a range of 32km and max speed of 20kts, with the sensors effective up to 6-8kts, with washout above 6kts. The operation is as follows. You must be at 4kts as before, and enter the presets in the same way. After firing the weapon it will begin to feed back data immediately and move at 4kts. The speed of the UUV is controlled with the enable button and the depth is controlled with the preenable button. The preenable button has no effect on the passive sensor. One click of the enable button will stop the UUV; it can persist indefinately in this state (although I will most likely have a timer on it in the full version), a second click will speed the UUV up to 6kts. A third click of the enable button will speed the UUV up to 12kts, the max speed the UUV can travel in up to 90ft of water without cavitating. A fourth click of the enable button and the UUV will go to its max speed of 20kts. A fifth click will stop the UUV and reset the counter, although you can click the enable button twice slowly and set it to 6kts. Note the sensors are washout above 8kts and do not feed data. The preenable button depth control works as follows. A single click does nothing. A second preenable click will send the UUV to the preset search depth. A third preenable click will send the UUV to 90ft if it is in over 100ft of water or 45ft if it is in less than 100ft of water. A fourth click will send the UUV back to launch depth, and reset the cycle.

SLAM-ER and Misc. Missiles:

The SLAM-ER now works for ASuW use and will enable a radar seeker at the last waypoint if it is over water. If the last waypoint is over land, the missile will operate in Strike mode, and behave as a light TLAM. The missile has a stealth enable feature that sends it down to just above the ocean before enabling and then after enabling it rising back to its cruising altitude of 30ft. Note, the standard harpoon has also been giving this cruising altitude, and the flight profiles of various missiles have been lowered. Also, the standard SLAM has been fixed and equipped on the AI P-3 as a land attack missile because the AI can't use the SLAM-ER properly. The AI P-3 does carry the Harpoon for ASuW. The Harpoon and the SLAM-ER both have a 40 Radar PSL, which is very low.

Helicopters:

I have attempted to fixed several problems like crashing and reporting contacts at launch, as well as dragging the active dipping sonar. Please use the FFG AI MH60 as well as observe helo behavior in general. :-) Also, the MH60 no longer launches with its radar on. :-)

CIWSAttack Doctrine has been updated to give better intercept performance will allowing for appropriate missile conservation for sustained attacks.

I made a minor change to the TLAM doctrine to make sure it always explode near the target as opposed to disappearing if it overshoots.

The Random Direction Torpedo Mod has been disabled for all torpedoes to allow more predictable subroc and AI MH60 weapon delivery. The torpedoes will always go to the right upon enabling.

The Hull array of the SW has been changed to simulate what we believe to be more close to the actual sonar suite on the SW. The Hull array on the SeaWolf now represents a low frequency receiver with coverage slightly larger than the Sphere array and with the same geometry. The frequency sensitivity and and washout speed remain as before. This sonar suite should be very helpful for tracking evading targets and as well as for all situaions in the littorals where the TA cannot operate.

The Maxspeed of the MPT torpedo payload on the SS-N-27 ASW has been reduced from 55kts to 45kts. This is done partially because the torpedo probably is closer to that speed as well as to reduce the effectiveness of the SS-N-27 relative to the new torpedoes behavior.

That's it!

Please play the heck out of this, specifically looking into these things I have mentioned here and provide as much feedback as you can through the usual channels. :-)

Cheers,
David
LW

Please do not post this file to any other site, as it is NOT an official LWAMI distribution. Thank you.

Cheers,
David

And, of course, I forgot to mention the "obvious", the ADCAP and UGST will run slower when deep. :)

LW, so 1500-3000 UUV sensor works for you ecactly the same like 0-2000 ? On my comp difference is so drastic like between detecting targets 10-20nm out just after launch, and detecting nothing at all initially, only if distance reduced to less than 2-3nm.

And if you mean there is no difference between 0-2000Hz +5 sonar and for example 1500-3000Hz -5 sonar, then I didn't checked this precisely, but for other sensors there was a clear difference (low frequencies propagate far, high freq sonars are semi-capped in range, even very loud contacts are muted at longer ranges), and I think I have seen the difference for UUV sensor too - the proportion between loud target det ranges and quiet target det ranges was quite different.

For example, low freq sonar that detects Akula II 5kts from 300yd, detects noisy civilian from 20+ nm

high freq sonar tuned to detect Akula II 5kts from 300yds, detects noisy civilian from 5-7 nm only. Thats the difference (of course numbers are pure example) And it could be even greater if there were more frequency absorbtion bands above 1500Hz in sonar model... :-/ But now seems (in DWAnalyzer) that highest frequency that matters is 1500Hz... have to check if it's the same in game, there is no passive sensors set to lower limit above 1500Hz in original DB, but there are active sensors with much higher freqs.

P.S. I wonder if we are using same game version to tests and if not maybe that's why we see different things :lol: I have single install of the game only, if it's telling you something ;)

I tried the version of the game you are using a bit... and nothing.

I am really getting no effect whatsoever from the frequency range.

Is there any data available (or about to be available after you get some rest?) on the effects of the ATP?

I was thinking something along the lines of a chart with x=speed, y=range with a few representative curves for different depths or something like that. Really, anything just to give us a ballpark idea how it's actually going to work. =)

Hi,

You may be making things more difficult for yourselves than need be. The sonar sensor model is a very simplified one. It only cares about one frequency. For this discussion, I will create a vessel profile with freq1 = 50, freq2 = 125, freq3 = 320, freq4 = 1050, freq5 = 1500. Now lets create 2 sensors, one is a LF sensor with a freq. range of 10-800. The other is a MF sensor with a freq range of 1000-3000.

When using the LF sonar, the only freq. of interest is freq3. (320). If freq3 happens to be 810 instead of 320, the LF sensor will not see the vessel. Freq1 & 2 play no roll except to add individual lines to the narrow band display.

The MF sonar only cares about freq5. If the line at freq5 is outside of the MF range, again, the vessel will not be seen. So changing freq5 to 3500, will make the sensor useless.

If we create a sensor with a range of 10-3000, the only freq. of interest is freq5, (1500Hz).

At least this is how the model worked for SC and DW1.01. I put the game away after that.

This simplified sensor model was created in order to save cpu cycles. The game only has to worry about one discreet freq. instead of many.

cheers, jsteed

Quote:

Originally Posted by jsteed

Hi,

The MF sonar only cares about freq5. If the line at freq5 is outside of the MF range, again, the vessel will not be seen. So changing freq5 to 3500, will make the sensor useless.

It's indeed very simplified, what have striken me lately is the extremely crude frequency absorbtion modeling, well at least if it works like Ludger's tool shows.

I have made even 8000-20000 and 19000-20000 AI sensors for UUV and they did detect contacts. As there are no discrete lines at all that high, they really should be useless in game but they worked, probably exactly same like a 1500-3000 sensor. I will reply te tests to be absolutely sure, probably will try them on torpedos too.

P.S. From what freq up a passive sonar is treated by game as an MF passive ? Or is there a separate flag for it in DB maybe ?

Instead of making LW or Ami do this for us, I went into the doctrines myself, and after much head scratching and fist pounding, came up with this... I think this is simple enough to do in combat if you have a calculator handy. =)

Range, Variable with Speed
ADCAP Range = 27 - 6(SetSpeed - 40)/15
UGST Range = 27 - 6(SetSpeed - 35)/15
(short version, every 5 knot decrease increases the range by 2nm, from a base range of 21nm)

Range, Variable with Depth
ADCAP Range (and speed) reduction = .29(depth{feet})/3000
UGST Range (and speed) reduction = .2(depth{meters})/735

LW, Ami, if this is wrong, please correct me.

Not sure if this is a problem with the new mod version or the way I installed it...
I've noticed two problems with the FFG controls for the Helo (when AI helo is controled by the FFG)

1) Whenever I assign a waypoint to the helo I find that the helo takes a while to start dipping. It circles around a bit around the waypoint and then drops down to dip.

2) This is a bigger problem, sometimes if I assign a waypoint to the helo while it's dipping, it ignores the new waypoint. It's the damnest thing. Assign a waypoint, it does nothing, delete it and assign a new one, it rises and moves a bit but then goes back to dip at the same spot. (it doesn't matter what kind of waypoint it is - Fly to, buoy, torpedo) This problem comes and goes. I could play a mission for an hour or so and then this happens. (I checked the usual suspects. The helo is within range of my FFG, It's in Sync)

Like I said, I'm not sure if it's just my install or a bug. Thought you should know...