Visual Measurement of Very Long Range Distances

[ColonelSandersLite]

Before continuring, I'd like to say that thoughts are welcome. If you have a better way, please let me know.

Ok, so the problem is: how does one get distance measurements on a distant target *BEFORE RADAR* is available. This can be pretty important when trying to set up an attack on an escorted convoy for example.

Alright, here's what I've come up with, after some reading, critical thinking, and math. In a nutshell, you can measure the distance to a target at very long range using the horizontal scale on the periscope and tbt. Using this method, I can get measurements that are reasonably accurate (+/- about 200 yards) on a target that is 5 miles away.


As we all (probably) know, the standard method up close is to use mast height and either the stadimeter or vertical hash marks as can be seen here:
http://img398.imageshack.us/img398/2958/plateiiawg1.jpg

The problem with this method at long range is that you cannot necessarily see the waterline of a ship at long range due to the curvature of the earth (which is modeled in sh4) so I got to thinking, what if you use the horizontal hashes instead?

The steps to follow:

1: You must be at or near a 90 degree aob from the target, so shadow the target and pull even with it.

2: Measure the targets length using the horizontal hash marks.

3: Reference ship length and do some math.

The following is what I have come up with for tmo 2.5. Depending on how the game displays, it may be relevent that I'm using a 16X9 Resolution.


Periscope - High Zoom TBT

277.63*Ship Length 515.58*Ship Length
Distance = ------------------ Distance = ------------------
Hash Marks Hash Marks

Ship Name Length - Yards Ship Name Length - Yards

Japanese DDs & Escorts-------- Merchants---------------------
Auxilary Subchaser 68.90 AKITA MARU 114.83
Akikaze Destroyer 112.10 BIYO MARU 131.23
Akizuki Destroyer 147.64 British Med Old Tanker 100.07
Asashio Destroyer 129.05 BUZYUN MARU 112.64
Etorofu Escort 72.18 Coastal Comp Freighter 88.33
Fubuki Destroyer 129.59 Conte Verde Liner 190.00
Kisaragi Destroyer 111.55 HAKUSIKA MARU 147.64
Minekaze Destroyer 112.20 HARUNA MARU 83.11
Momi Destroyer 112.20 HEITO MARU 113.30
Mutsuki Destroyer 111.55 Hog Island typ A Frghtr 131.23
Shiratsuyu Destroyer 118.11 HORAI MARU 150.33
Somers Destroyer 126.86 KASAGISAN MARU 94.93
Type C Escort 97.88 KINPOSAN MARU 118.11
Type D Escort 97.88 KITURIN MARU 142.17
Shimushu Escort 76.55 Landing Ship Tank 107.94
Wakatake Destroyer 112.20 Large German Tanker 207.79
Yugumo Destroyer 129.05 Liberty Cargo 160.76
Medium Modern Passenger 114.83
Japanese Carriers------------- MOMOYAMA MARU 131.23
Akitsu Escort Carrier 157.21 NAGARA MARU 149.82
Chitose Seaplane Tender 210.52 NIPPON MARU 164.04
Hiryu Fleet Carrier 244.20 Old Raked Bow Split 131.23
Shokaku Fleet Carrier 273.40 RYUUN MARU 131.23
Taiho Fleet Carrier 277.45 Small Old Split Frghtr 88.04
Taiyo Escort Carrier 196.96 Small Split Freighter 95.00
T3 Tanker 207.79
Japanese BBs and Cruisers----- TAIHOSAN MARU 88.04
Agano Light Cruiser 191.38 Troop Transport 166.23
Furutaka Heavy Cruiser 205.60 TYOHEI MARU 86.40
Fuso Battleship 231.85 Victory Cargo 152.67
Ise Battleship 240.59 ZINBU MARU 131.23
Kongo Battleship 240.59
Kuma Light Cruiser 178.26 Japanese Others---------------
Maya Heavy Cruiser 223.10 Atami Gunboat 49.21
Mogami Heavy Cruiser 218.72 Hira Gunboat 60.15
Naka Light Cruiser 177.16 Okinoshima Minelayer 134.95
Takao Heavy Cruiser 223.10
Yamato Battleship 287.62

If you're not using TMO, the ship lengths can be gotten with http://www.subsim.com/radioroom/downloads.php?do=file&id=1602

I took these steps to determine the constant.

Set up a single mission with a couple of stationary vessels.

Start the mission with map contacts set to on, so that you can measure the distance to target on the map. Radar could be used instead.

Count the number of hash marks the target takes up and plug the data into this formula.

Constant to use = (Distance X Hash Marks) / Ship Length


Recommendations:
When deriving the constant, close to the point where the sample vessel takes up all the hash marks. Try to avoid fractions (IE 8.5 hash marks).

Test it on a few different vessels, not just one. I can't say with any certainty that all the measurements the game spits out are actually correct.

Make sure you're at 90 degrees AOB.

Test it at a few different ranges to make sure you got it right.

If you're plotting in yards, be sure that the ship lengths are converted from meters to yards. The game's values are metric so Gutted's utility will also spit out metric values.

[ColonelSandersLite]

Been doing some reading on the OTC mod, and it seems that the above formulas are *only* valid at 1920X1080 due to the way the game displays.

I was also thinking about it and, while I haven't tried it, you should be able to skip converting to imperial lengths to plot in yards if you determine the constant to use from metric lengths. The conversion would be built into the constant then.

[TorpX]

If I am reading your post right, you are using the length of the target instead of the height, to estimate the range. I don't see this as practical, because you will not be able to rely on having the target at an AoB of 90 deg., and without map contacts, will not be able to accurately estimate the AoB. This would add a additional source of error, and would be less reliable than just using the stadimeter and some sort of correction for the hull down effect.

If you set up your test with ships only at a 90 deg. AoB, you might get good estimates, but how could you replicate this in a realistic scenario?

In RL, they just had to accept the limitations of the stadimeter until RADAR came along. Skillful plotting and "fairing the plot" helped. In any case I do not consider range estimates beyond 5,000 yds. or so, to be very reliable in this game.

[ColonelSandersLite]

In any case I do not consider range estimates beyond 5,000 yds. or so, to be very reliable in this game.


This is exactly the point. It's a method to get some sort of range estimate beyond 5k yards, and my experimentation shows that it works pretty well out to 10 thousand or so. Basically as long as you can make out the bow and stern of the ship, so effective varies a bit with the size of the ship.

What you don't need is an exact 90 degrees AOB. works well +/- about 10-15 degrees, which is easy to estimate imo.

I know that the numbers I'm saying in this post are kind of vague, but that's because I don't have enough practice at it yet to be more specific.

Obviously though, pre-radar, I agree that what awe can do is certainly limited precision wise. I don't know about you though, but my stadimeter readings in the 6k yards+ range are just complete garbage regardless of attempts to compensate.

[TorpX]

I don't know about you though, but my stadimeter readings in the 6k yards+ range are just complete garbage regardless of attempts to compensate.
Same here.

In RL, I think the reason the vertical dimension was used was because the ship would always be vertical, unless it was already listing. Using the length means adjusting for the AoB, which is undesirable. The way ships are rendered graphically, though, seems to create bigger problems for us. I'm speaking of the way the masts seem to grow or shrink, from moment to moment. (I guess everyone has this problem?) Is there any chance of eliminating this?

Maybe you got something here. :hmm2:

[ColonelSandersLite]

Compensating for AOB only needs to be done to the same extent that AOB needs to be compensated for when using the speed formula. Many people will state that the speed formula gives reasonably accurate results +/- about 15 degrees, so as long as it looks like it's in the ballpark, it's probably ok.

I'm speaking of the way the masts seem to grow or shrink, from moment to moment. (I guess everyone has this problem?) Is there any chance of eliminating this?

This effect is a result of screen resolution combined with the size of the polygon used at that specific lod. In a nutshell, if the mast width is less than 1 pixel wide, depending on the way the ship is bobbing and swaying, and the way the graphics software handles rounding the value, parts of the mast will not be rendered.

Increasing the screen res will help up to the point where you hit the maximum. Anti-aliasing can help or hinder depending on the exact algorithm used and the amount of AA applied.

At the end of the day though, current technology isn't even close to handling the resolution of the human eye and that's where the problem lies.


For anybody interested:

I currently run a 23" 16X9 Monitor at 1080p

Viewed from 20" away, the resolution required to match the "standard 20-20" human eye would be something like 9000X5893

This amounts to actual image output being about 4% of the ideal

At 32bit color (true color), this is 1697184000 bits, which is 202.32 megabytes *per frame*.

[ColonelSandersLite]

Forgot to mention, that yes, I would agree that the advantages of using mast height far outweigh trying to use length when closer than about 6000 yards since the apparent mast height will not change very much with aob. My method is only intended to give us a tool to use in a range bracket where the stadimeter is totally useless.

[TorpX]

Is there any chance of eliminating this?



Viewed from 20" away, the resolution required to match the "standard 20-20" human eye would be something like 9000X5893

....

I'll take that as a no. :-?

[ColonelSandersLite]

This technique is now battle tested and works way better than I could have hoped for.


http://i3.photobucket.com/albums/y91/ColonelSandersLite/sh42013-01-0422-26-29-93_zpsc088fac3.png

The situation:

I was playing the photo recon single mission. In this mission, I did not have radar. I had already completed the photo recon part of the mission and there where 2 destroyers ahead of me about 5,000 yards distant, and a destroyer behind me about 4,000 yards distant, all of which where looking for me and royally ****** off. When I came up to periscope depth, I saw a Nippon maru in the distance, which was my last objective.

I quickly determined that a collision course intercept was out of the question, the angles where just wrong and I had to stay at silent running if I wanted to get my fish off. This meant that I had to head right towards the firing position and risk a long range salvo. Since I did not have radar, the best tool in my arsenal was measuring his length as I've been discussing in this thread.

As you can see, looking at the plot, my first two readings (marks 23 and 24) where off. I attribute this to his AOB just being too sharp. It should be further apparent that measurements from mark 26 onwards where right on the money, with an average error of maybe 25-50 yards. Considering that every one of these marks where visually measured at a range in excess of 5,000 yards, I consider that margin of error to be outstanding. Examining mark 26 in further detail, I can tell you that his AOB at that mark was 38 degrees and his distance was 11,200 yards.

The end result of this track was 4 mark 14's fired from the bow on fast setting at a range of about 4,600 yards. If you know your mark 14s, you know that this is out of range, except that this is his range *when you fire*. He actually closed the last hundred yards in time for two of my torpedoes to hit him, slowing him to a crawl so I could finish him off later at my leisure.

[TorpX]

That's certainly a good stalking there.

I still wish there was a way to use the mast ht. of the ships, though.