MaGui 3.4 Widescreen AoB finder problem

[GOREUNCLE]

Hi everyone.

Since all the MaGui related threads I've found are closed, I have to start this one, sorry.

So I've been using MaGui 3.4 Widescreen since it was released (GWX Gold installation), I just love this mod, specially the manual targeting bit.

The thing is that I started using the "perfect shots" method by Markman94 and I noticed that the gyroangle solutions I was getting were incorrect.
My torpedoes used to pass right in front of the target's bow, meaning that my gyroangles (or shooting bearings, as it turned out) were too wide, so I started investigating this issue.
I quadruple-checked my "lb" angle measurements, I quadruple-checked (and even more, XD) that the formula (SB=G+x) was applicable to the situations I was in, I quadruple-checked (and more) my speed estimation accuracy, I read all the relevant documentation dozens of times, etc.

My "lb" angle measurements were correct, the shooting bearing formula was correct, my speed estimations were very accurate, according to the documentation I was doing everything correctly...only one thing remained in question:
The "x" angle reading from the AoB finder.

So here's a picture of the x12 AoB finder that comes with MaGui Widescreen 3.4:


As you can see, the red "x" mark is pointing at a location way to the left of that "90" value (to its right).


And here's a picture of the standalone MaGui x15 AoB finder:


As you can see, that red "x" mark is pointing directly above the "90" value.


So I wondered "what if the corrections made in the x12 AoB finder for MaGui Widescreen, didn't require that red "x" mark to be relocated to the left?".
Eureka!!! That was exactly the issue. The moment I started using the "x" angle values located directly above that "90" in the x12 AoB finder, my shooting bearings became perfect... the "perfect shots" method started living up to its name.

So here's my modified x12 AoB finder, which gives correct "x" angle values:




I'm hoping that Markman94 will read this and provide his input on this issue (I really think there was no need to relocate that red "x" mark in the widescreen version. Since the discs were already adapted, that "x" mark should've remained in the same relative position... At least that's the explanation I came up with).
And, if it's ok for him, I'm posting a video tutorial I made. This tutorial shows how the "perfect shots" method works (in the most typical scenario). Make sure your browser allows Youtube video annotations, since I've included some.
Ideally, you don't want to use torpedo salvo offset with this method (because it alters the interception angles of the torpedoes), but I did in this video.
Also, this video shows a short range shot, so the discrepancy of 1.4º in "x" is negligible, but it wouldn't be negligible at longer distances:



I hope you find this post useful, I know this discovery worked wonders for me.

Thanks to Markman94 for coming up with this manual targeting method and sharing it with us, it's my favorite method by far.

[makman94]

hello GOREUNCLE,

very nice video and comments

about the 'x' mark:
the 'x' mark is placed correctly on the MaGui's discs. the problem is somewhere else and i will try to explain it.



the 'x' angle is the angle (of a right triangle) opposite the side of target's course. the formula for this angle is: tan(x)=u1/u2 <=> x=arctan(u1/u2) where u1=target speed and u2=torpedo speed (use this formula for creating your tables that you are mentiong at your video.the tables that you are showing there are completely wrong and you must redo them.These tables are allready done for you as you will find out reading lower)


When discs are on their starting positions the '30' for a 30kts torpedo is matched with the '30' for a 3kts target (the 40kts torpedo is matched with a 4kts target and the 44kts torpedo is matched with a 4,4kts target).

3/30=4/40=4.4/44=tan(x) <=> tan(x)=1/10 <=> x=5,71°

so,when discs are on their starting positions the 'x' mark must point at the ''5,71'' of the inner marks of middle disc. As you can see the x mark is correctly placed to all setups for discs (either on the x15 setup or either on the x12 setup the 'x' mark is pointing at the ''5,71'' of middle disc)

so where is the problem ?
At low speeds of target there is no problem but ,as the target's speed increase, there is indead a fault reading for 'x' angle on the tool. This fault is not constant but it is increasing following a logarithmic fuction. in other words , the faster the target is,the bigger the fault that discs are giving as angle 'x'.

this is not caused by the 'placing' of 'x' mark. this issue is caused becuase the inner marks of middle disc are the tangent (or sine) scale for small angles (<5,71°). the discs are using the ''ST'' scale there which is a scale for tangent (and sine) of small angles. this scale is perfect for use at ranges (and not only) but it is not the proper one for use in order to give correct outputs for the 'x' angle.
The 'x' angle needs the ''T'' scale ,which is the scale for angles >5,71°,in order to get accurate readings from the discs.in other words, one more scale (the 'T' scale) must be added in the middle disc in order to read from there the 'x' angle.

i hadn't noticed this issue when i released the method back then becuase i used in my testing senario a low speed target (5kts) and i got a correct reading from disc(at low speeds works fine as i said before) so i moved on. Later, i discovered that as target's speed increase so is increasing the fault at 'x' angle reading and i found the cause of the issue that i mentioned above.

of course ,adding this scale to the discs ,i don't believe that worth the effort for two reasons:
1.the discs are allready 'heavy' numbered and
2.the 'x' angle can be found very fast from somewhere else.The angle 'x' is allready calculated in the shooting tables that come with the MaGui so you can use these tables for getting the ''x'' angle. i had edited ''The broken TDC Attack Method'' thread and showed there how to get the angle 'x' from these tables.
(http://www.subsim.com/radioroom/showthread.php?t=169935)

ps: this method is not taking in account the parallax correction (the difference at bearings between torpedo tube 'eye' and periscope 'eye') so expect some tiny inaccuracies there.the further the target is,the less the parallax correction is needed.so, Avoid using this method for very close range attacks and prefer shootings to >1500m

all the best,
Manos

[GOREUNCLE]

Originally Posted by makman94

hello GOREUNCLE,

very nice video and comments

about the 'x' mark:
the 'x' mark is placed correctly on the MaGui's discs. the problem is somewhere else and i will try to explain it.



the 'x' angle is the angle (of a right triangle) opposite the side of target's course. the formula for this angle is: tan(x)=u1/u2 <=> x=arctan(u1/u2) where u1=target speed and u2=torpedo speed (use this formula for creating your tables that you are mentiong at your video.the tables that you are showing there are completely wrong and you must redo them.These tables are allready done for you as you will find out reading lower)


When discs are on their starting positions the '30' for a 30kts torpedo is matched with the '30' for a 3kts target (the 40kts torpedo is matched with a 4kts target and the 44kts torpedo is matched with a 4,4kts target).

3/30=4/40=4.4/44=tan(x) <=> tan(x)=1/10 <=> x=5,71°

so,when discs are on their starting positions the 'x' mark must point at the ''5,71'' of the inner marks of middle disc. As you can see the x mark is correctly placed to all setups for discs (either on the x15 setup or either on the x12 setup the 'x' mark is pointing at the ''5,71'' of middle disc)

so where is the problem ?
At low speeds of target there is no problem but ,as the target's speed increase, there is indead a fault reading for 'x' angle on the tool. This fault is not constant but it is increasing following a logarithmic fuction. in other words , the faster the target is,the bigger the fault that discs are giving as angle 'x'.

this is not caused by the 'placing' of 'x' mark. this issue is caused becuase the inner marks of middle disc are the tangent (or sine) scale for small angles (<5,71°). the discs are using the ''ST'' scale there which is a scale for tangent (and sine) of small angles. this scale is perfect for use at ranges (and not only) but it is not the proper one for use in order to give correct outputs for the 'x' angle.
The 'x' angle needs the ''T'' scale ,which is the scale for angles >5,71°,in order to get accurate readings from the discs.in other words, one more scale (the 'T' scale) must be added in the middle disc in order to read from there the 'x' angle.

i hadn't noticed this issue when i released the method back then becuase i used in my testing senario a low speed target (5kts) and i got a correct reading from disc(at low speeds works fine as i said before) so i moved on. Later, i discovered that as target's speed increase so is increasing the fault at 'x' angle reading and i found the cause of the issue that i mentioned above.

of course ,adding this scale to the discs ,i don't believe that worth the effort for two reasons:
1.the discs are allready 'heavy' numbered and
2.the 'x' angle can be found very fast from somewhere else.The angle 'x' is allready calculated in the shooting tables that come with the MaGui so you can use these tables for getting the ''x'' angle. i had edited ''The broken TDC Attack Method'' thread and showed there how to get the angle 'x' from these tables.
(http://www.subsim.com/radioroom/showthread.php?t=169935)

ps: this method is not taking in account the parallax correction (the difference at bearings between torpedo tube 'eye' and periscope 'eye') so expect some tiny inaccuracies there.the further the target is,the less the parallax correction is needed.so, Avoid using this method for very close range attacks and prefer shootings to >1500m

all the best,
Manos

Wow, thanks a bunch for this great reply!
I'll definitely dig into the info you provided, I'll use those MaGui tables (I always thought they were only meant to be used with the straight shots method, but I see now that I was wrong).

I'll perform further tests using the "x" values provided in those tables and see how it goes.

I'll keep you posted.

[makman94]

Originally Posted by GOREUNCLE

Wow, thanks a bunch for this great reply!
I'll definitely dig into the info you provided, I'll use those MaGui tables (I always thought they were only meant to be used with the straight shots method, but I see now that I was wrong).

I'll perform further tests using the "x" values provided in those tables and see how it goes.

I'll keep you posted.

the angle 'x' belongs to a right triangle on which the one side is the torpedo's course run and the other is the target's course run.
so,the 90° columns of these tables are simulating perfectly this situation.

for your tests , i will suggest you to create a test mission with your mission editor on which you will know exactly the target's speed and course.

have fun,
HTH

[GOREUNCLE]

...I just wrote a loooong post about a test I made using the MaGui "x" values...with conclusions, possible things that could be going on, etc...
The post got deleted after I tried to post it because my login expired (why didn't I save the post like I always do? XD)

Well, long story short: the MaGui "x" values from those tables don't work.
They produce the same result as the placement of the red "x" mark to the left of the "90" value on the AoB finder. My torpedoes miss, passing right in front of the target's bow... the shooting bearing they give is too wide.

On the other hand, my custom "x" value hit the target...however, it missed the precise spot I aimed for by 2-5 meters (behind it, so that means the "x" angle should've been slightly higher, but not nearly as high as the MaGui's "x" value).

I performed the test in a way that both shooting bearings could be verified within seconds from each other (by firing two electrical torpedoes).
The first torpedo was launched at "MaGui's shooting bearing", the second torpedo was launched a few seconds after, when the target reached my "custom shooting bearing". As I said, MaGui's value was off by a few degrees, my custom value was off by less than a degree.

Tomorrow I will perform further testing to make sure this isn't an isolated result, if it keeps happening, I'll post the results and possible explanations I see.

[GOREUNCLE]

Well, I performed further testing.

4 patrols with a total tonnage sunk of around 300.000 tons.
I performed lots of attacks of all kinds (as you can imagine), in all of them, my modified "x" values turned out to be correct.

The supposedly correct "x" values from the MaGui tables and AoB finder produce incorrect shooting bearings (too wide). As I said before, my torpedoes end up passing right in front of the target's bow.

I know this shouldn't be the case, but it is. I mean, when I use my custom "x" values, I end up hitting the target exactly at the spot I aimed for (even at ranges of 1500m or more), but when I use the MaGui/AoB finder values, I miss.

I don't know what's going on here, but I can say that my speed estimations and lb measurements are performed correctly...

It would be great if someone else could test this, to find out whether or not there's a problem with my SH3+GWX+MaGui widescreen installation which creates this weird behavior.

[makman94]

Originally Posted by GOREUNCLE

Well, I performed further testing.

4 patrols with a total tonnage sunk of around 300.000 tons.
I performed lots of attacks of all kinds (as you can imagine), in all of them, my modified "x" values turned out to be correct.

The supposedly correct "x" values from the MaGui tables and AoB finder produce incorrect shooting bearings (too wide). As I said before, my torpedoes end up passing right in front of the target's bow.

I know this shouldn't be the case, but it is. I mean, when I use my custom "x" values, I end up hitting the target exactly at the spot I aimed for (even at ranges of 1500m or more), but when I use the MaGui/AoB finder values, I miss.

I don't know what's going on here, but I can say that my speed estimations and lb measurements are performed correctly...

It would be great if someone else could test this, to find out whether or not there's a problem with my SH3+GWX+MaGui widescreen installation which creates this weird behavior.

hello GOREUNCLE,

there are many many reasons for a torp to pass in front (or back) of target and certainly the answer is not at the angle 'x'. as i told you above, the 'x' angle is the angle (of a right triangle) opposite the side of target's course so it is the trigonometry (and only the trigonometry) which will give you the angle 'x' values. you can't make 'custom' tables ,thats a 'no go' route.

somewhere else is hiding the answer and this is the not accurate data for target's speed and course.even at single test missions ,you set a target doing ,for example, 5 kts and in game ,for an unknown reason, i see the target doing 4 kts. this is not happening for every ship though, other ships are doing 5 kts.
Other tip is that you must set the gyroangle clockwise! for example, if you want to set the gyro at 20° you must do it clockwise.that way you are sure that gyro is at 20°. if you set it at 20° counter-clockwise then the gyro is ,in fact, setted at 21° which will lead to faults ,especially at long shots.
Also, make sure that you are ,as possible, exact at your shooting bearing.(even 0,5° at long shots is critical)


anyway, i am trying to say this:
When you want to test your tools you have to make single missions that you know the exact target's data and see the results. also, at your test ,use ONLY electric 30kts torps(the target's manouvers for avoiding the steam torps are amazing...and surely you can't make safe conclusions).

Originally Posted by GOREUNCLE

.... but when I use the MaGui/AoB finder values, I miss.

I explained why not to use the RAOBF rings for getting the angle 'x'.i have discard the 'x' mark on the RAOBF rings at my next uis for the reason that i mentioned above. you must use the ''90° column'' of shooting tables for getting the angle 'x'


here,try the following mission:
http://speedy.sh/SftFU/Test-A-Perfect-Shot.rar

it is jsgme ready,drop it in your mods folder and run the game.
you will see a mission called ''Test A Perfect Shot'',in this mission the target is doing exactly 18 kts ,heading to exactly 270° (Impact range is at 4800m).
shoot it with a 30kts torp and tell me your results

edit: here is a video with my shot at the above test mission:

[GOREUNCLE]

Hi Markman.

I've been working with your mission and with a modified version I've made, the results I'm getting are mixed. The MaGui values work in some cases (long range), but the shooting bearings they provide are too wide in other cases (medium-close range), making the torpedoes miss.

My custom values worked in some cases and missed in some others.
So I started looking into stuff that could explain this. At first I thought this had to be related to range, target's lenght, etc. But today I've discovered something disturbing: the ruler tool is broken.

I've discovered that the distance values given by the ruler tool can't be trusted, here's why:


If I can't get reliable distance data, I can't calculate the target's speed using the 3:14 method, thus, my shooting bearings will be wrong, regardless of the "x" values I decide to use.

Tomorrow I'll keep performing tests and try to figure out what's going on with the ruler tool...

EDIT: Ok, I just realized that the ruler tool actually rounds the values up. The distance shown above is actually 350 meters, now it all makes sense.
The first line is actually 350 meters long.
The line configuration below it consists of two 150 meter lines and one 50 meter line.
The last line configuration is made of 7 lines of 50 meters. All the line configurations measure a total of 350.

So every map square equals 50 meters (at full zoom), put two map squares together and that's 100 meters... Now I can finally get accurate distance/range readings without even having to use the ruler tool, XD.
Great! Now that I know this, I can perform further tests with accurate speed readings.

[makman94]

Originally Posted by GOREUNCLE

Hi Markman.

I've been working with your mission and with a modified version I've made, the results I'm getting are mixed. The MaGui values work in some cases (long range), but the shooting bearings they provide are too wide in other cases (medium-close range), making the torpedoes miss.

My custom values worked in some cases and missed in some others.
So I started looking into stuff that could explain this. At first I thought this had to be related to range, target's lenght, etc. But today I've discovered something disturbing: the ruler tool is broken.

I've discovered that the distance values given by the ruler tool can't be trusted, here's why:


If I can't get reliable distance data, I can't calculate the target's speed using the 3:14 method, thus, my shooting bearings will be wrong, regardless of the "x" values I decide to use.

Tomorrow I'll keep performing tests and try to figure out what's going on with the ruler tool...

EDIT: Ok, I just realized that the ruler tool actually rounds the values up. The distance shown above is actually 350 meters, now it all makes sense.
The first line is actually 350 meters long.
The line configuration below it consists of two 150 meter lines and one 50 meter line.
The last line configuration is made of 7 lines of 50 meters. All the line configurations measure a total of 350.

So every map square equals 50 meters (at full zoom), put two map squares together and that's 100 meters... Now I can finally get accurate distance/range readings without even having to use the ruler tool, XD.
Great! Now that I know this, I can perform further tests with accurate speed readings.

hello GOREUNCLE,

thats why i am telling you to make your tests at testing missions where you are sure of the exact target's course and speed.so ,use the single mission that i sent you ,create in there your senarios by changing target's speed,course and uboat's course and make your tests.Once you are sure for target's speed and course,you move on to:

The parallax correction is getting ''wilder'' as closer the target is and as bigger the gyroangle is.
So, avoid using this method for close targets or with extreme gyros (>1500m and +-20° gyroangles will be fine).
The shooting tables are 100% correct and these are the tables that you must use in any case.

for accurate distance measurements, the MaGui is using a very very usefull tool, the Malloy Ruler.
Search into the documents for the tools of MaGui and study the tool. with Malloy Ruler you can get 100% accurate distance measurements at any zoom level of map. for start ,practice yourself with the tool at the 0-500m of map scale


all the best

[GOREUNCLE]

Ok, so I think this case can be closed now.
The reason why I couldn't get accurate shooting bearings with either the AoB finder and the MaGui tables was a combination of parallax correction and incorrect distance measurements (therefore incorrect speed estimations).

I was about to go crazy because the MaGui table values still didn't work at ranges of 1000-1500 meters even with the correct speed estimations, but it seems this is caused by the parallax correction you mentioned.

That's why my custom "x" values worked, they compensated for the parallax correction error in a way, and the wrong speed estimation compensated for my custom "x" values being wrong (all of this without knowing it, XD).

I've found that the MaGUi table values can also be used at ranges between 1000-1500 meters, but a degree must be subtracted from them to compensate for the parallax correction (I'll need further tests to confirm this, but so far it has worked very well).

As I said yesterday, I realized that the map scales are there for a reason! .
It's very easy to measure distances just by counting the number of squares traveled by the target in a given amount of time. I just need to check the scale of the map in a given zoom level and then I'll now how many meters does each square represent.

I have the readme for the Malloy tool, but I don't find the MooseMeasure folder... I'll see if I can find the files somewhere.


Thanks for your assistance Markman94, without your input I doubt I would've solved this puzzle.

Now I guess I'll need to make a new video, correcting all the mistakes I made.

Btw, I want to ask a couple of things:

1- In the video that you linked, I saw that the Bismark had a very nice damage model. What visual and damage model mods are you using?

2- Do you know where can I find the files for the Malloy tool? I'll look in the downloads section, see if it's there.

Thanks a bunch!

[GOREUNCLE]

Thanks for the link mate!

I'm having a problem, though. I can't extract the file, both winrar and 7zip give me an error...

[GOREUNCLE]

Originally Posted by Plissken_04

I've updated the Link,so try it again!


So Long

Maik

I'm sorry to say that it gives me the same error as yesterday "unexpected end of data". Can you upload the file somewhere else? Perhaps there's a problem when downloading files from the location you are linking...

[makman94]

Originally Posted by GOREUNCLE

....

I have the readme for the Malloy tool, but I don't find the MooseMeasure folder... I'll see if I can find the files somewhere.

......

Btw, I want to ask a couple of things:

1- In the video that you linked, I saw that the Bismark had a very nice damage model. What visual and damage model mods are you using?

2- Do you know where can I find the files for the Malloy tool? I'll look in the downloads section, see if it's there.

Thanks a bunch!

hello GOREUNCLE,

the Malloy ruler is allready build in the files of MaGui (see the ruler how big it looks ingame). Once you understand how it works ,you will be able to take very accurate distance measurements.

about your questions:
1. it is just a texture image from a mod here:
http://www.subsim.com/radioroom/down...do=file&id=962

2.Plissken's storage place is the best to look at when you are aftering for mod(s). as i said, the Malloy Ruler is allready in the game so you don't have to install anything. download the mod from Plissken's place just for its documents folder (Malloy had put in there some explaining pics)