Spain and Seaplanes

[pagsab]

Yes, I noticed the famous forts from the classic siege described in Ernle Bradford's book.  SF2 terrain making is always a balance between the desire for accuracy through bespoke tiles and TOD objects for individual features and the wish to keep tile and TOD variants as few and manageable as possible.  I lean towards the latter priority, to the perhaps ridiculous extent of using just one tree clump model and two building models across my entire tileset!  I keep these small to enhance the sense of speed when flying, since I find that some otherwise beautiful terrains have trees so big that low-flying planes seem to be moving in slow motion.  Realistically sized troops and vehicles are very difficult to see in time to line up for a strafing run, but contrast helps a lot.  One trick that I just discovered is to darken the texture file of vehicles to stop the camouflage working so well that they blend in completely with the ground.   I limit shadows to maintain frame rates and avoid distracting shadow 'popping', so darkening target objects offers a useful substitute.  Conversely, it is important to keep the contrast between TOD terrain objects and the tile background as low as possible so that one notices less when the objects pop into or out of viewing range.  

[pagsab]

I just finished making the MTO equivalent of my reduced scale Straits of Dover terrain as showcased in https://combatace.com/forums/topic/101179-european-air-war-reborn/.  As seen below, my latest effort has the same 45 by 32 tile layout, though it portrays a larger area of real terrain since each tile represents 5.1 km instead of the 3.3 km scale which I use in crowded NW Europe.  Both terrains saw constant air combat across the straits from June 1940, with the Allies gradually gaining the upper hand until they finally ended the stand-off by amphibious invasion 3 or 4 years later.  I used the same tiles and modelling techniques as in my SCW terrain, though with forests and scrubland generated only on tiles with an elevation range of 120 metres or more.  The towns and cities represent all those with 30,000 or more inhabitants in the 1936 census (as shown in Wikimedia historical population data).  Using rescaled HFDs and procedural tile selection made the process much quicker than my initial entirely manual efforts, and I have already begun retro-fitting these techniques to my NW Europe project.  I can now fly complete Allied and Axis combat missions in both theatres, and an obvious next step is to expand my modelled areas to London and Catania respectively to give a broader range of bases and targets.  With my SCW terrain also crying out for expansion, and with my current reading of White's gripping 1986 book on 'The Gotha Summer' constantly tempting me to create a WW1 variant of my evolving North Sea terrain, I am spoilt for choice as to which avenue to pursue first!

[pagsab]

My SCW and MTO projects have reinforced the importance of combining online and specialist hard copy research sources – neither alone is sufficient to provide the nuanced details one needs.  I am increasingly concerned that my used copy of Herrera's SCW seaplane book has still not arrived from Madrid, but Howson's big 1990 Putnam reference work on 'Aircraft of the Spanish Civil War' did just get to me in Chile at last after a 2 month journey.  Eagerly browsing its seaplane entries, I realised that my static Lohner conversion is wasted portraying the old Macchi M.18, which performed only liaison and training duties during the SCW.  I hence souped up my M.18 conversion as shown below by boosting its speed and engine power and adding a rear gunner using the PilotBalkans2B figure.  The resulting model looks much more like the Savoia-Marchetti S.62 than did my earlier stand-in of the Supermarine Walrus, despite sacrificing 3D guns, a spinning prop and visible damage.  The sturdier Walrus will make a better stand-in for the Curtiss and Felixstowe flying boats in my planned WW1 North Sea project.  I am meanwhile enjoying exploring the Sicilian coast in my favourite seaplane of all – the Regia Marina's single seat Macchi M.41 which I brought to life from Stephen1918's other static model.  It is a shame that this 'Riva Aquarama of the skies' was replaced in 1938 by the ugly and fatally flawed IMAM Ro.44 floatplane.  Howson's book also details how three Savoia S.55s (the type famous from Italo Balbo's mass long distance flights) played a minor role in the opening months of the SCW, but finding a stand-in for a big twin hull flying boat is asking too much even for me!

[pagsab]

Terrain makers face interesting choices over whether and how to modify the natural height of their elevation points.  Some terrains deliberately increase real heights to create more dramatic relief, but in the mountainous countries of Spain and Italy I have had to halve the actual heights to match my doubling of the horizontal distances represented by each tile, and so stop slopes becoming unrealistically steep.  This generally works well, as shown by the picture below of a Macchi 202 over my newly completed model of NE Sicily (with the Etna massif just visible on the distant horizon).  However, in small rugged islands like the Lipari isles shown here, the paucity of elevation points makes the jaggedness of slopes look jarring, and to soften the effect I have opted to halve heights yet again in these islands (with still greater reductions for single tile islands like Stromboli in the foreground where even a single elevation point raises some of the surrounding water).  Moving to the high definition alternative of 64 rather than 16 elevation points per tile would help with this small island problem, but at the unacceptable cost of requiring far too much editing to make coastal elevations fit with my chosen tiles.  This is yet another of the challenging trade-offs which make SF2 terrain creation an art as much as a science.  (Plane modelling raises similar dilemmas, and I have now decided to add the GunBarrel and GunBarrelRear models from the pilot folder to my Savoia S.62 shown above – far from ideal because of their excessive calibre, but better than the flat painted lines which I used hitherto.)

[gerwin]

"However, in small rugged islands like the Lipari isles shown here, the paucity of elevation points makes the jaggedness of slopes look jarring"

The solution is to contract the height values away from the coast a little, and let the coast relief be determined by per tile _HM heightmap files. See my Persian Gulf and IcelandNA terrain for examples on how to do that. IcelandNA has steeper _HM files compared to Persian Gulf.
Same goes for rivers.

[pagsab]

Thanks - that's a fascinating alternative.  Wrench said in my FE2 thread at https://combatace.com/forums/topic/101534-opening-the-adriatic/ that a proper set of water tiles needed these heightmap files.  I have dispensed with them so far in my own tilesets for the sake of simplicity, but I will need to explore their potential as you suggest.  One apparent downside is that every instance of a given tile will use the same attached heightmap unless I am mistaken.  I use my small island tiles in multiple locations (including for peninsulas like the Milazzo peninsula on the right of the picture below), and it is good to be able to choose their height in each individual case.  However, it may be worth trying to make a couple of distinct rugged single tile islands using your method to overcome the difficulty I described.  Meanwhile, I just finished modelling the Italian toe, and celebrated with a late afternoon flight from Reggio in a CR-42.  The picture shows the classic view across the strait to Messina, with the arc from Etna to the Lipari isles in the background.  It is wonderful to be able to modify and retile existing HFD files to create such vistas.  Now I just need to model NW Sicily to have the island as a whole.

   

 

[Wrench]

That looks really good!

now, here comes the joke ---

will the Italian bunion be modeled as well??

.... I'll show myself out ....

[pagsab]

You are excused, for having been first to remind me of the significance of heightmap files for water tiles!    After my usual blundering experimentation, I finally got Gerwin's suggested method to work, and as soon as I get time, I plan to try out various ideas with this way of adding high definition terrain relief.  I will report back in due course.

[pagsab]

I have reported in my new thread at https://combatace.com/forums/topic/101610-detailed-terrain-relief the great success I have had applying Gerwin's suggested use of heightmap files while completing my Sicily and Malta terrain.  However, a question there about seaplane modelling prompted me to conduct extensive additional experimentation to see just what can and cannot be done in SF2 with these classic craft.  Since this topic fits better with the current thread, I will outline my findings here for ease of reference.

My Spain and Sicily terrains show that it is easy to create seaplane bases by commenting out the ModelName in the relevant Runway entry in the terrain's Types.ini, just as for grass airfields.  The problem is that most seaplanes which start at such offshore bases move at only a crawl if at all, and so fail to take off for their missions.  This is especially annoying since the same seaplanes anachronistically take off perfectly well from airfields on land, while landplanes take off perfectly well from the sea!  Seeing that Spitfloat models did manage to take off from the water prompted me to see if power to weight ratios were the key.  Sure enough, other seaplanes also took off after I cut their mass in their data.inis by 50 or even 75 percent, but this is hardly a practical solution given its drastic effect on the planes' overall performance.  At correct weights, SF2 makes water so treacle-like that, even at full power, most seaplanes never reach flying speed.

One way around this is to give the seaplanes wheels as well so that they gain the anachronistic benefits enjoyed by landplanes.  Amphibians like the Walrus and Catalina operate well from the water as long as one remembers to raise the wheels after take-off and drop them for landing.  Stephen1918 created a few small seaplanes for FE2 which have both fixed wheels and floats, and these too seem to operate well from the water even in SF2.  Unfortunately, their success seems to depend on their tailored LODs, and all my attempts to add 'virtual' wheels to other seaplanes by data.ini changes alone caused the models to explode immediately unless I added 'NoContactonGround=TRUE' to the phantom wheels, in which case they were predictably useless.  Main Gear entries in the Landing Gears section of the data.ini must be either wheels or skids, with the latter distinguished by 'IsSkid=TRUE'.  Skids expose planes to the torpor I have mentioned in the water, but have the compensatory benefit that one may amend their 'Contact Point' entry to raise the model above the water on a cushion of air until explosions due to collisions cease to occur (especially important with the conversions which I have described earlier in this thread).  Wheels ignore ContactPoint entries and rely instead on LOD data, which is useless if the wheels are non-existent or sit above any added floats.

A second very simple way of getting around the treacle-like water is to drop one's Landing setting to Easy in gameplay options.  This allows one's own plane to take off normally at the outset or after alighting on the water, although overweight or underpowered craft can still struggle.  Unfortunately, I have not found a way to shift AI planes to a similar Easy setting, so the rest of one's flight or squadron is left behind as one soars above in frustration.  A key reason for my own interest in seaplanes is that the trick I use to fly as a wingman in SF2 (see https://combatace.com/forums/topic/100072-following-your-leader-through-the-low-european-clouds) means that my colleagues begin already in the air, waiting for me to take off and join them.  Players who prefer to lead their flights will benefit less from changing the Landing setting, though since seaplanes flew alone more often than did landplanes, the trick still offers useful benefits for complete first person sorties.

A final way of escaping the problem is to avoid having seaplanes start on the water.  The easiest way of doing so is to start them instead at a nearby land base in the in-game mission editor.  A more satisfactory alternative for one's own flight or squadron is to edit the saved mission file so that the aircraft mission concerned reads 'StartOnGround=FALSE' and the first two entries are converted to standard Waypoints with the initial height raised from 0 to 500 metres or so.  You should now begin flying over the seaplane base, having missed the thrill of take-off but ready to lead your assembled colleagues to the mission objective.  Since AI planes disappear straight after landing, there is no problem in having the mission end at a seaplane base, as long as you raise Contact Points enough to avoid landing damage, and as long as you make Landings easy and enable Tailwheel steering in the data.ini if you wish to taxi back across the water to the hangars.

I really hoped that my concentrated search would discover a killer workaround which would make seaplane missions as immersive and unconstrained as those of landplanes.  However, even the techniques I have mentioned combine to yield remarkable results considering that seaplane modelling was never even thought of in the original Third Wire sims.  If anyone knows of any additional tricks one can use, please do share them here.

[Wrench]

if using (indvisibile) wheels, why not raise the RollingRadius= to keep them of the ground?

It's obvious you've access to the DAT sea & floatplanes. I personally haven't tested the PBY from YAP's Rising Sun, so can't comment on it. And as OUR Catalina is -5A, with wheels, it's never been an issue.

(oth, it sure would be nice to have some static RA and Luftwaffe birds to 'park' at seaplane bases!)

[pagsab]

Modifying RollingRadius is a fascinating idea, and does indeed have a similar effect on wheeled planes to changing ContactPoint Z values for skids.  I even got an Arado 95W into the air with all its skids changed to wheels in the data.ini, at the ridiculous expense of the floats spinning like tops.  However, my wingman remained stubbornly mired in the treacle until he finally crawled up onto the land, removing the point of the whole Frankenstein exercise.  Meanwhile, an unmodified pair of wheeled Arado 95As took off from the water with contemptuous ease.  So frustrating.  I am glad that my own preference for wingman flights means that my existing tricks usually suffice to enjoy a full seaplane experience like the Macchi 41 formation flight around Sicily which I describe in my new thread.   The picture below as we passed Capo San Vito shows the scenic benefits of my new incorporation of heightmap files as per Gerwin's suggestion.