Second attempt QGIS

Post new mods and scenarios here.

Moderator: MOD_Flashpoint

exsonic01
Posts: 1133
Joined: Tue Jul 26, 2016 6:45 pm
Location: Somewhere deep in appalachian valley in PA

Second attempt QGIS

Post by exsonic01 »

Now I'm trying my second attempt of QGIS mapmaking after epic fail. [X(] This time, I will 99% follow the manual's progress.

Image

Image

So far so good

This time I used proper hex size setting (433m x 500m) and proper EPSG reference (WGS 84 UTMzone52N). My current AO is 25km x 30km sized, contains 57.5 (in zig-zag direction count, x-dir) * 60 (linear count, y-dir) hex cells. If I regard 57.5 as 58, there are total 3480 hex tiles.
exsonic01
Posts: 1133
Joined: Tue Jul 26, 2016 6:45 pm
Location: Somewhere deep in appalachian valley in PA

RE: Second attempt QGIS

Post by exsonic01 »

Image

Print composer work is a new pain. I used 650mm x 750mm page size, keeping scale 39370. But, of course, the hexes are off from FCRS map editor hexes. Working in the loop of save and load and adjust min/max extent.... Hopefully, wist to find the good match within few tries.

Is it better to work with roads / river earlier?
User avatar
WildCatNL
Posts: 929
Joined: Fri Mar 27, 2009 5:21 pm
Location: Eindhoven, the Netherlands

RE: Second attempt QGIS

Post by WildCatNL »

Looks good, the numbers make sense.

One suggestion: why not work out a small section of the map first in full detail, before going into "mass production". I did that for the 90x60 hex Eiterfeld map, and it helped my sanity.
You could either export a subsection of the map, for example as a 20x15 map, and do that one first. Or just create a Print Composer for the full map (something the guide doesn't cover) and leave a lot empty.

William
William
On Target Simulations LLC
User avatar
WildCatNL
Posts: 929
Joined: Fri Mar 27, 2009 5:21 pm
Location: Eindhoven, the Netherlands

RE: Second attempt QGIS

Post by WildCatNL »

ORIGINAL: exsonic01
Print composer work is a new pain. I used 650mm x 750mm page size, keeping scale 39370. But, of course, the hexes are off from FCRS map editor hexes. Working in the loop of save and load and adjust min/max extent.... Hopefully, wist to find the good match within few tries.

Is it better to work with roads / river earlier?

@PrintComposer: yes, they are a pain to create. Since you have a large map, the bottom right corner hex might not fully line up with FCRS hex grid when the top left one does. If so, bumping up the scale slighty, say from 39370 to ~39385, might fix this (once the top left hex lines up).

You have a river in your map, so start with the river. See the updated guide.

William
William
On Target Simulations LLC
exsonic01
Posts: 1133
Joined: Tue Jul 26, 2016 6:45 pm
Location: Somewhere deep in appalachian valley in PA

RE: Second attempt QGIS

Post by exsonic01 »

ORIGINAL: W1ll14m

Looks good, the numbers make sense.

One suggestion: why not work out a small section of the map first in full detail, before going into "mass production". I did that for the 90x60 hex Eiterfeld map, and it helped my sanity.
You could either export a subsection of the map, for example as a 20x15 map, and do that one first. Or just create a Print Composer for the full map (something the guide doesn't cover) and leave a lot empty.

William

The region will probably be the last line of defense right before the city of Pyongyang. Low mountains and hills from east to west in the middle of the map provide a sort of natural defense line to defenders. Right above (5km north) the map, city of Pyongyang begins. So defenders will fight desperately.

I was considering to depict the battle for three stages in single scenario:
- approaching to the defense line,
- battle for the defense line, and
- breakthrough to the north, get the crossroads and threat the city itself.
Naturally, battlefield should be longer in y-dir, needed to be 30km to simulate all three phase of the battle. Is working part by part would better? Then, how can I add each parts and align them? Reason why I didn't do part by part is, I just wished to finish the map in one go.

Or, if the 25x30 is not a good idea, I will cut east 5km and south 5km, that will give me 20km x 25km. Is this looks good enough?
ORIGINAL: W1ll14m
ORIGINAL: exsonic01
Print composer work is a new pain. I used 650mm x 750mm page size, keeping scale 39370. But, of course, the hexes are off from FCRS map editor hexes. Working in the loop of save and load and adjust min/max extent.... Hopefully, wist to find the good match within few tries.

Is it better to work with roads / river earlier?

@PrintComposer: yes, they are a pain to create. Since you have a large map, the bottom right corner hex might not fully line up with FCRS hex grid when the top left one does. If so, bumping up the scale slighty, say from 39370 to ~39385, might fix this (once the top left hex lines up).

You have a river in your map, so start with the river. See the updated guide.

William

Ok, I will start working on the river and roads after print composer works. Seems like I need to have perfect match of hexes before go to rivers and others... and I need to think about map size.
exsonic01
Posts: 1133
Joined: Tue Jul 26, 2016 6:45 pm
Location: Somewhere deep in appalachian valley in PA

RE: Second attempt QGIS

Post by exsonic01 »

Following Wiliam's recommendation, I made a new smaller area box, 20km x 25km, 46 x 50 hexes. Delete east 5km and south 5km from previous 25 x 30 square km map.

Image

Image

Image

New smaller map takes out some room for flanking maneuver. Purpose of previous larger map was to offer a chance of flanking enemy defense lines from unexpected direction. Oh well, at least this would be easier for me to play with. In addition, the new map still contains two important major roads, hills and mountains, and numerous channels, which satisfies my plan for the scenario.

Print composer page size = 529mm x 661mm, 39370 scale. Current setting gives me 2665pixel x 3331pixel under 128 dpi condition.

Image
Now working in the loop for the hex match. Looks like my hex need to move 10% right, 45% up
User avatar
BeirutDude
Posts: 2811
Joined: Sat Apr 27, 2013 9:44 am
Location: Jacksonville, FL, USA

RE: Second attempt QGIS

Post by BeirutDude »

Looks like we're in the same place! [:D] [&o] [8D]
"Some people spend an entire lifetime wondering if they made a difference. The Marines don't have that problem."
PRESIDENT RONALD REAGAN, 1985

I was Navy, but Assigned TAD to the 24th MAU Hq in Beirut. By far the finest period of my service!
exsonic01
Posts: 1133
Joined: Tue Jul 26, 2016 6:45 pm
Location: Somewhere deep in appalachian valley in PA

RE: Second attempt QGIS

Post by exsonic01 »

My 3 month old boy fortunately sleeps well tonight :) Finally, got a time to play with map.

Hex match achieved. Followings are four corner hexes of my map. Target hex was the one in top left corner. But as you can see, top right, bottom left, and bottom right hexes are very slightly off (like 5m~10m) from the FCRS hex positions. This will potentially create headache in bigger map, and I think it is what William mentioned in the manual about the problem of the big maps. (My current map (20km x 25km) is slightly bigger than recommend size (20km x 15km))

So if anyone plans even bigger map, it would be better to check all four corners carefully. Also, it would be better to make the map part by part (but I'm not sure how could I do that, specially matching hex at the boundary part)

Image

Image

Image

Image

Created .fp9 using automated map scanning option for all layers. Used default option (3 pixel edge buffer). All hex's visual hindrance are 4, and mobility hindrance are 5 (Of course, I did nothing on the map other than topo data). But elevation looks OK. Not '99' in any layers.

Image

Willam, is this looking good? Should I need to send the current fp9 and png or later with roads, waters, and woods?

Also, may I have some questions?

1) I matched hexes from QGIS with FCRS using QGIS' print composer like manual did. Meanwhile, the hexes in QGIS window (not print composer window) would still might be in off position from FCRS's hex position. Is this potentially could be the problem when I try roads, rivers, and city based on rendered image and fp9 file in future?

2) I want to cut out the edge area out of while area-box line. Should I need to use photoshop or paint?

3) If I wish to put grid reference (ruler image) at the edge of the map, is this available inside the QGIS?

4) Can I modify the thickness of hex and 1km grid line in later stage?

Thanks

Best,
exsonic01
Posts: 1133
Joined: Tue Jul 26, 2016 6:45 pm
Location: Somewhere deep in appalachian valley in PA

RE: Second attempt QGIS

Post by exsonic01 »

ORIGINAL: Beachinnole

Looks like we're in the same place! [:D] [&o] [8D]


Your works are faaaantastic~!
User avatar
BeirutDude
Posts: 2811
Joined: Sat Apr 27, 2013 9:44 am
Location: Jacksonville, FL, USA

RE: Second attempt QGIS

Post by BeirutDude »

William can answer this better than I but here's my thoughts...
I matched hexes from QGIS with FCRS using QGIS' print composer like manual did. Meanwhile, the hexes in QGIS window (not print composer window) would still might be in off position from FCRS's hex position. Is this potentially could be the problem when I try roads, rivers, and city based on rendered image and fp9 file in future?

Right now rivers and streams are the hex boundary issue and that is set in the scanner manually. Keep your roads to the center of the hexside and as much as possible perpendicular to it. Snap your urban/village locations to the hexside vertices and fill the entire hex. No partial towns and stick to the big ones (I was doing farms and small industrial sites as well). So there will be compromises where the towns/roads/rivers will be slightly off it's real location or larger than reality if you really want it in. Map making has always involved some compromise, and wargame maps in particular. Have to learn to suppress that inner perfectionist tendency we all have as gamers! [:D]
I want to cut out the edge area out of while area-box line. Should I need to use photoshop or paint?

Neither, go back onto QGIS and turn that layer off, then go to print composer and there us a map preview button. Hit that and it will update your layers in composer.
If I wish to put grid reference (ruler image) at the edge of the map, is this available inside the QGIS?

I've seen that option in composer, but why? Each hex is approximate 500m so that provides scale in its own right. If it were me I'd skip this but I believe it's in the drop down menus if you want. One other thing might mess up the scanner? Remember the goal is to get the map in the scanner not map a finished map (see below).
Can I modify the thickness of hex and 1km grid line in later stage?


Again why? The scanner is set for a hex thickness and the QGIS map I learned is a draft and William has a tool that makes the pretty maps from the QGIS FP9 file> So I've learned don't do work you don't need to do. Two other things. I reduced the thickness of the river to help the scanner out with mobility. and set the scanner buffer a bit higher at Jim's suggestion. So if your rivers are coming out as lakes those are two things you can do to help the scanner.
Your works are faaaantastic~!

Thank you but william is to be credited for all of his help, I would never have gotten this far without him. There are a lot of things I learned and the next maps will be better! [8D]
[:)]
"Some people spend an entire lifetime wondering if they made a difference. The Marines don't have that problem."
PRESIDENT RONALD REAGAN, 1985

I was Navy, but Assigned TAD to the 24th MAU Hq in Beirut. By far the finest period of my service!
exsonic01
Posts: 1133
Joined: Tue Jul 26, 2016 6:45 pm
Location: Somewhere deep in appalachian valley in PA

RE: Second attempt QGIS

Post by exsonic01 »

Thanks Beachinnole. Answeres helped me a lot. I wish to put reference ruler image at the edge, and 5km (or 1km) rectangular grid to the map, to help players to judge the distance during game. Regarding question 1, I will think about that later. I'm worried that I may need to re-adjust hex match later after I finish roads, forests, citys, waters, and etc, but I think I can do that without any serious problem.

Thank you~! [&o][:)][:D]
exsonic01
Posts: 1133
Joined: Tue Jul 26, 2016 6:45 pm
Location: Somewhere deep in appalachian valley in PA

RE: Second attempt QGIS

Post by exsonic01 »

Got a one more question.

Image
Google map

Image
OCM

Image
OSM

I decided to use OCM map as a reference of water, road, city and all others, and ESRI Satelite will be used for the double checking. Regarding forest, I may use U.of.Maryland data but that will be in the future. Google map / Bing map is slightly off from other maps in current UTM 52N coord., and ESRI maps are unable to magnify from certain scale limit. For this region, I think OCM/OSM are best option to play with.

But as you can see, OCM map has great detail, specially in road network. Do I need to draw all those roads? Some of those white lanes in the map might be unpaved, 1~2 lane roads, or even smaller, and drawing all of them would take lots of time. Road condition and quality of NK is quite messy, so there would be the road which might not proper our game, I'm not sure which one to choose and not.... Or, I could use OSM, to depict large-white-lanes only, and ignore small-white-lanes.... Are there any criteria to ignore/draw roads?






User avatar
BeirutDude
Posts: 2811
Joined: Sat Apr 27, 2013 9:44 am
Location: Jacksonville, FL, USA

RE: Second attempt QGIS

Post by BeirutDude »

I would only draw "military important roads". I put that in quotes as that may change from area to area. In an area with a dense road network only major highways might be important as all of the county roads male the other hexes the same. Conversely in the Sinai a dirt track may be military significant as it is better than the rocky, rough terrain around it. So drawing county roads in New Jersey is futile but county roads in Arizona more important. I hope you see what I'm getting at.
"Some people spend an entire lifetime wondering if they made a difference. The Marines don't have that problem."
PRESIDENT RONALD REAGAN, 1985

I was Navy, but Assigned TAD to the 24th MAU Hq in Beirut. By far the finest period of my service!
exsonic01
Posts: 1133
Joined: Tue Jul 26, 2016 6:45 pm
Location: Somewhere deep in appalachian valley in PA

RE: Second attempt QGIS

Post by exsonic01 »

ORIGINAL: Beachinnole

I would only draw "military important roads". I put that in quotes as that may change from area to area. In an area with a dense road network only major highways might be important as all of the county roads male the other hexes the same. Conversely in the Sinai a dirt track may be military significant as it is better than the rocky, rough terrain around it. So drawing county roads in New Jersey is futile but county roads in Arizona more important. I hope you see what I'm getting at.

I clearly understand what you mean, and I agree with you. The roads of North Korea are mostly unpaved, narrow ones. Usually it is hard to expect the full mobility on such conditions. In addition, it is fact that too much detail would cost too much time and concentration, so I think I would rather ignore them... mostly. Thanks
exsonic01
Posts: 1133
Joined: Tue Jul 26, 2016 6:45 pm
Location: Somewhere deep in appalachian valley in PA

RE: Second attempt QGIS

Post by exsonic01 »

Image
That stream has averagely 30m~40m width. So, that entire hex should be modeled as major river? Just to check, because, to me, it feels bit strange to approximate the entire 500m sized hex to water, only because of 35m wide river.


Image
One more example of tricky river... This stream has 30m~50m width, but I'm afraid that modelling it as major river would bring complicated map. Maybe I need to just put more approximation and regard this stream as minor river, to reduce number of major river sections, but I'm not sure this is good idea or not...
User avatar
WildCatNL
Posts: 929
Joined: Fri Mar 27, 2009 5:21 pm
Location: Eindhoven, the Netherlands

RE: Second attempt QGIS

Post by WildCatNL »

ORIGINAL: exsonic01
Willam, is this looking good? Should I need to send the current fp9 and png or later with roads, waters, and woods?
Looks good. Just send me what you have, and I'll have a go at rendering it. On my side, it is a bit of time to render it the first time. Any subsequent updates from you won't make time.
1) I matched hexes from QGIS with FCRS using QGIS' print composer like manual did. Meanwhile, the hexes in QGIS window (not print composer window) would still might be in off position from FCRS's hex position. Is this potentially could be the problem when I try roads, rivers, and city based on rendered image and fp9 file in future?
Both of them should show the same alignment for the features you draw on the map. The hex grid may seem to shift a bit wrt Google / Bing maps "underlays" when QGis isn't switching to "Pseudo Mercator" coordinate system. (As a rule, I draw on top of the underlays using Pseudo Mercator, but always export in UTM).
In the end, the only thing that counts is the print composer.
2) I want to cut out the edge area out of while area-box line. Should I need to use photoshop or paint?
As beachinnole say, just disable the layer and export the bitmap using the Print Composer.
3) If I wish to put grid reference (ruler image) at the edge of the map, is this available inside the QGIS?
I can add these to the map I render for you. I don't do that in QGis.
4) Can I modify the thickness of hex and 1km grid line in later stage?
.
Yes. Treat the problem as having to create two maps: the first to feed the "game" (Map Values scanner) and create the .fp9 map file. Once that's done, you can create a prettier variant any way you like (QGis, Paint/Photoshop) and display that in game.
[What roads to draw?]
Eyeballing probably works best for Korea, unless the area features a system where the road designation indicates the road's importance (as they have in Germany).
Some suggestions:
- Keep in mind that urban hexes in FCRS are given the same mobility (hindrance) value as road hexes, so you do not need to replicate dense urban networks
- Anything that has a vehicle capable bridge attached is significant. That's why all my maps feature rail roads...
- Roads that in the real world aren't offering mobility advantages to tracked / four wheel drive vehicles can be ignored.
[Expand a 35m wide water obstacle into a 500m hex?]
All maps serve a purpose, and abstract the world and deviate from a 1:1 scale for that purpose. Roads are drawn wider on McNally road atlasses / Michelin maps than they are in real life to aid the reader.
From the purpose of the game's user, a commander in his HQ, there is a significant difference between water obstacles that can be cleared by a standard bridge layer and water obstacles that require a pontoon bridge or amphibious capability. Especially if your force lacks amphibious capability but your opponent has it...
Note also that in game terms, there is little downside to shifting and compressing hexes around the wide water obstacle, since it will be the same for NATO and WP.
As a guideline, if deviating from the "real" (paper/satellite) map makes the in-game combat better reflect the Cold War fighting that might have occurred, just deviate.

Again, don't hesistate to send your work to me. It would be great to have a first Korean Theater map for the Flashpoint Campaigns game!

William
William
On Target Simulations LLC
exsonic01
Posts: 1133
Joined: Tue Jul 26, 2016 6:45 pm
Location: Somewhere deep in appalachian valley in PA

RE: Second attempt QGIS

Post by exsonic01 »

Thank you so much for all those answers and helps William, I sent fp9 and png files. Regarding the river, OK, I will try to use large river hex tile to any stream which has 20m+ width.
exsonic01
Posts: 1133
Joined: Tue Jul 26, 2016 6:45 pm
Location: Somewhere deep in appalachian valley in PA

RE: Second attempt QGIS

Post by exsonic01 »

Facing some weird problem in "snapping" option for minor river ( = major stream)

Image

As you can see, QGIS snapping option is recognizing the corner of hexes off from what it appears. All of layers are in same coordinate (WTS52N) but not sure why this happens. Should I need to shift the hex grid? I just did large rivers, so shifting them would not that difficult...

ps) I re-made the hex grid, and now it looks fine, and finished the print composer work again. Sorry William, I send you new files, please ignore files in previous emails. Just use ones from the most recent email.

Reason of this symptom is still unknown. But it would be always better to check the hex grid snapping time to time.
exsonic01
Posts: 1133
Joined: Tue Jul 26, 2016 6:45 pm
Location: Somewhere deep in appalachian valley in PA

RE: Second attempt QGIS

Post by exsonic01 »

Image

Image

Image

Image

Finished water + Highway + some of the primary roads. AO is full of streams, rivers, small~medium towns, trenches, fortifications, bunkers and etc... looks like it would be a tough map for attackers.
exsonic01
Posts: 1133
Joined: Tue Jul 26, 2016 6:45 pm
Location: Somewhere deep in appalachian valley in PA

RE: Second attempt QGIS

Post by exsonic01 »

Finished all roads and all river/stream networks. "Digitized" the road more carefully, to perfectly fit in the hex grid system.

Image
How it looks like in the topo map

Image
South west corner of the AO, OSM data

Image
South west corner of the AO, Satellite data

To draw the roads, I mainly used 70% OSM and 25% OCM and 5% ESRI satellite data. Picked secondary roads only when it has certain width or more. Now next step is draw banks, towns, fields, woods, and etc... Current style is FCSS styles, and I will convert them in FCRS style later.
Post Reply

Return to “Mods and Scenarios”