PDA

View Full Version : Deadly accurate stadimeter rangefinding.


Neutrino
03-18-09, 07:18 AM
I only started playing SH3 a short while ago but one of the things I noticed was that using the stadimeter I appeared to be getting consistantly long readings for the range on some ships, errors in the order of several hundred meters at 5km, I think I have now found the cause of that.

The ship recognition manual and game config files list the highest point on the ship as the mast height. I'm not sure this is authentically accurate. I was under the impression that the mast height on the ship is supposed to be the height of the tallest mast proper, that is not includings flags, penant holders, antenna or whatever else might be hanging off the mast. I can't find a reference for this atm so if anyone knows for sure please pipe up.

The reason for this would be obvious. Thin antenna, pennant holders and the like will be invisible in low visibility.

So what I have been doing is taking a range measurement at the top of the mast head, which at long range or in poor visibility is likely to be the highest point on the ship you can see, and the notebook has been using a different point on the ship that I can't see as the reference point, and that's been the cause of my error.

I've had a think about it to try and work out what might be the best way around this. I can measure the actual height of the mast head quite easily. I think I've found the files containing the ship recognition data and the data used by the stadimeter. It seems to me that the following could be one possible solution:

1. Measure the height of the mast head.
2. Update the ship recognition and ship config data with that new height.
3. Update the picture of the ship in the recognition manual to include a faint red line to indicate at which point the height reading was taken.

I'd be interested to hear what the rest of you think of that.

1. Has it already been done?
2. Is it likely to be as straightforward as I suspect?
3. Is there a better way?

One final consideration.

Makman94 has mentioned he has a mod which uses funnel height to achieve a similar end. I have a few concerns about that approach, but I'm not sure how valid they are so your views on that would also be appreciated.

In situtations where the target presents a small aspect to you, ie you're almost directly ahead or astern of it you may have trouble identifying the top of the funnel. It may be obscured by superstructure. Large cargo ships can have 4 or more cranes, the ones in front of the funnel will always obscure it, in darkness where all you see is a silhouete the masts both in front of and behind the funnel will blot it out. One of the main advantages of stadimetric rangefinding is that it works no matter what angle the target presents to you. For the reason just stated using the funnel as a reference point risks undoing that advantage.

The advantage of using the funnel is that in many cases it is more centrally located than the heighest mast, which for a large ship presenting a small aspect gives you a fix closer to the center of the ship, which should be more accurate than getting a fix on a mast near the bow or the stern.

I can forsee a couple of problems with this.

Firstly the advantage of getting a fix on the center of target is only really useful if the ship is presenting a low aspect to you, (if it's broadside on a range at any point along it is good). But it's precisely when the target presents a small aspect that getting a fix on the funnel is likely to be more difficult.

Secondly. The funnel is likely to be significantly lower than the tallest mast. For stadimetric rangefinding the higher the reference point the more accurate the reading. So by using the funnel as a reference point you risk getting a less accurate result to start with.

And one final point.

If you're taking a range fix on a large ship that is presenting a small aspect or travelling directly away from you, the smart captain will know that if the ship is 100 meters long and he's taking a fix on a mast near to one end of it he should add 30 meters or however much from his calculation as appropriate. In this manner smart play and experience provide the rewards of a better outcome, which I like.

In conclusion: This is unlikely to be something I'll be able to knock out in a week or two, I have a lot of RL coming up. But I might make a start on it for my own satisfaction, and before I embark on that I just wanted to check with you guys that I haven't overlooked any important considerations.

Thanks.

makman94
03-18-09, 07:57 AM
One final consideration.

Makman94 has mentioned he has a mod which uses funnel height to achieve a similar end. I have a few concerns about that approach, but I'm not sure how valid they are so your views on that would also be appreciated.



hello Neutrino,i think that you don't understand exactly. my mod is using the mast heights (NOT the funnels heights) .did you try it?

ps:about the funnels.

if you don't run the funnel heights file then you play your game with the masts.but if you run the funnel heights ,you must remember that this file containes ONLY values for the five ships you will see at torpedo training.if you run this file ,i said that you enter ONLY the torpedo training and play your game but looking at funnels,not masts. this file is only for testing purposes ,just to see how funnels work.
did you try it?
don't worry about the target aspect.when i say 'looking at funnel' i mean the point that is at the middle top of funnel.funnels are much easier to be spoted when target is far away.and if target has 3 or 4 funnels (it will be marked at the rec manual to which one funnel to look)

ps2: remember that no matter how exactly correct the optics are setted...when the target is far,for example at 4 km then one pixel up or one pixel down at stadimeter is represending with 200-300 m (that depends from mast height also) and nothing can be done with that ! (as far i know). what you can do when target is far is: take one measurement up and one down.then calculate the middle value of these two measurments and you are closer to the real value.
and something else:no matter how high is the point you are looking for (mast or funnel).the false is the same.

good luck to your efforts,

bye

Neutrino
03-18-09, 05:41 PM
I understand that the funnel height experiment was just for one training mission. No I haven't tried that but I have tried getting fixes on ships funnels in the dark from ahead and astern and it wasn't easy which is one of the reasons why I said I wasn't keen on that approach.

When you said that 'my mod is using the mast heights' do you mean that in your mod you've already modified the ship recognition manual to use the correct masthead height for all or some ships? I uninstalled OLC for now and did my testing with the stock (GLX3) stadi, it was just easier because in stock you can pause the sim while taking readings, and tbh I prefer the GLX ui.

In a previous post you said 'also i had adjust the angular angle in order the stadimeter to work correct'. What did you do here? Did you make the stadimeter more accurate by more closely replicating the correct formula, or did you tweak the stadimeter formula with a constant in order to make it get the range more accurately without having to change the ship files, or something else? What I really want to know here is, if I use the stock SH3 stadi to more accurately measure the masthead height, will that counteract the improvements in your stadimeter or will it compliment them?

You also said that 'of course some tiny adjustments must be made at mast's heights but this is a huge work to be made by one'. But is it really that much work? I created a custom mission that contained all the merchant ships lined up in a row stationary in a dead calm, and then just cruised past them at 3000m and ranged them at different points using the stock stadi. That was how I discovered that the stock stadi uses the height of the heightest point on the ship rather than the masthead height. It wasn't that difficult and only took an hour or so, I could just as easily have calculated the correct masthead height and changed the ship recognition manual and config files. I'm not saying you're wrong but just wondering if there is something else you know that I have overlooked?

Madox58
03-18-09, 05:59 PM
In the relm of a Game?
You can make everything Perfect.

But during WWII at Sea, in a Real U-Boat?
How many things were perfect?

I prefer a simulation that somewhat matches those conditions.
Messed up recognition stuff, inaccurate information, etc.

In someways?
People add that questionable stuff to thier Mods.
They intend the miss information.

Correct it as you see fit.
It's your Game after all.

I prefer to deal with the questions in Game
rather then make everything a Turkey Shoot.

DeerHunter does that for me.

makman94
03-18-09, 08:25 PM
I understand that the funnel height experiment was just for one training mission. No I haven't tried that but I have tried getting fixes on ships funnels in the dark from ahead and astern and it wasn't easy which is one of the reasons why I said I wasn't keen on that approach.

When you said that 'my mod is using the mast heights' do you mean that in your mod you've already modified the ship recognition manual to use the correct masthead height for all or some ships?--no, i didn't touch the mast's heights of gwx 2.1 or gwx 3.0 BUT i setted different the stadimeter (angular angle)and if you touch this angle you must change the mast values also.thats why i said that mast heights need tiny fixes also(but as it is now is more acurate) I uninstalled OLC for now and did my testing with the stock (GLX3) stadi, it was just easier because in stock you can pause the sim(you can do that in olc gui also,press del key on numpad) while taking readings, and tbh I prefer the GLX ui.--remember that stock and gwx3 are different things at this point (stadimeter)...other angular angles,other mast values....

In a previous post you said 'also i had adjust the angular angle in order the stadimeter to work correct'. What did you do here? Did you make the stadimeter more accurate by more closely replicating the correct formula, or did you tweak the stadimeter formula with a constant in order to make it get the range more accurately without having to change the ship files, or something else? What I really want to know here is, if I use the stock SH3 stadi to more accurately measure the masthead height, will that counteract the improvements in your stadimeter or will it compliment them?

You also said that 'of course some tiny adjustments must be made at mast's heights but this is a huge work to be made by one'. But is it really that much work?--yes ,is huge work...not so simple I created a custom mission that contained all the merchant ships lined up in a row stationary in a dead calm, and then just cruised past them at 3000m and ranged them at different points using the stock stadi. That was how I discovered that the stock stadi uses the height of the heightest point on the ship rather than the masthead height.---how do you measure the EXACT REAL range? It wasn't that difficult and only took an hour or so, I could just as easily have calculated the correct masthead height and changed the ship recognition manual and config files. I'm not saying you're wrong but just wondering if there is something else you know that I have overlooked?

hmmm, a lot must be said now but i really believe that you must do your 'own' trip in that research first and then we will communicate easier.

Neutrino
03-19-09, 05:06 AM
That's fair enough Makman I'll look into it when I have time. I just wish I could work out what you meant by 'changing the stadimeter angular angle', working out what you've changed there from examining the mod's file only sounds like it might be the hardest part of the job.

I measured the exact real range by having the map run in God mode, in this way it automatically updates the map with the actual positions of any targets you can see. Then it is a simple case to measure the distance using Moose Malloy's ruler mod. Assuming the God mode map is putting the target in the correct position it is quite easy to range the target to within 10 meters, which is easily accurate enough to get a good calculation on the height of the masthead.

I'm off to Spain to see my old folks this week and then when I get back I'm off to the far east for some travelling. So it might be a while before I get back to this. But once again thanks for all your help Makman.

o7

makman94
03-19-09, 06:47 AM
Then it is a simple case to measure the distance using Moose Malloy's ruler mod. Assuming the God mode map is putting the target in the correct position it is quite easy to range the target to within 10 meters

just keep in mind that Moose Malloy's ruler mod gives acurate readings only when you are at 0-500m or 0-2,5km or 0-5km scales of navmap (don't use this tool on 0-250m scale)

bye