Monday, 6 February 2017

Maps repacked and moved to a new location

Since the Play Unreal redirect went down I've moved the Unreal Tournament 4 maps I'm working on back to MPF.

Multi Player Forums Redirect

Unfortunately while the MPF redirect is working for uploading and downloading with a web browser it is NOT WORKING as a redirect.

Most of my maps are back on the MPF redirect depository and in-game redirects are working again now.

DM-Landscape has not been uploaded because Epic broke something in the most recent build(?s) preventing the trees from displaying at longer distances. It almost seems like they hard coded foliage cull distance, changing it to 0.0 (used to disable distance culling) or 500,000.500,000 (very large values) in editor makes no difference.

The problem is long range sniping is what the gameplay on Landscape is mostly about. The trees have collision and function as cover... If you can't see the trees your shots get blocked by invisible collision. If you zoom, the trees do fade back in, but still, it's a problem.

If you're looking to chat with me about any of my maps (feedback, suggestions or w/e) I can be found on Discord.

UC Discord
Epic Discord

Saturday, 24 September 2016

DM-Codex Progress Report

My DM-Codex remake was getting a little too big (filesize) and I wanted to scale the whole map up a bit since it was starting to play a bit too tight after multiple in game changes to movement and player scale over several months.

I decided to redo the whole map from scratch, significantly reduced the filesize in the process and released that. Now I'm polishing it up and adding details. It's getting bigger again, but not as big as before, plus it's more detailed and looks better too.

Here are a few quick screenshots taken in the editor.

Note: There is something wrong with lighting builds in the current version of the editor. It seems like bounced lighting isn't being applied correctly and messing up the shadows. It'll look better once that's fixed (hopefully in the next version of the editor which rumour says is coming soon).

Wednesday, 17 August 2016

UT4 Editor - Using Blueprint Skewmesh

Among the more useful new features in the UT4 Editor are skewmesh blueprints. There are 3 types of skewmesh blueprints, BP_SkewMeshX, BP_SkewMeshY and BP_SkewMeshZ.

What these blueprints do is take a mesh and allow you to move one end of it in any direction (up or down and to either side) around one of the 3 axis (X, Y or Z). Skews are extremely handy for aligning trim, wall meshes and other modular meshes along angles they weren't designed to fit. They don't solve every mesh alignment problem but they do help a lot.

Anyhow like they say a picture is worth a thousand words so here are several thousand words worth.

This first is a BP_SkewMeshY with the default mesh (a Liandri assets pillar/column) assigned to it.

You can change the mesh your skewmesh is operating on in the details panel. Here I changed it to use one of the Liandri assets floor trim meshes.

And here's another an example of that mesh zoomed in close so you can see the seam where the beveled edges of two meshes (one skewed,  one not) meet. The top part looks a little funny but that's a problem with material alignment on that particular mesh, not with mesh alignments which are perfectly lined up with each other.

Here are two instances of a floor panel mesh. The one on the left is by itself (not skewed). The one on the right is another instance of the same mesh inside a BP_SkewMeshY. Notice the transform handle named "End" to the right of the skewmesh. That's the transform gidget we use to move that end of the mesh around. This is quite similar to selecting and moving surfaces with the BSP editing tools.

Next, I've pulled the mesh upwards to make a ramp and added another instance of the mesh to the right. Notice how the edges of the skewed mesh align perfectly with the other meshes... Unlike when trying to get the same look by rotating and/or scaling the same mesh. They also light better and have fewer problems with z-fighting too.

Finally, here's the same skewmesh skewed to 45 degrees to one side rather than up. If I'd wanted an angled ramp, I could have skewed it both up and to the side at the same time.

There are some problems with various materials on skew meshes (they look fine in editor but show the default checkerboard grid material in game) but you can work around that by merging meshes. You can merge just a single skewed mesh or multiple meshes (both skewed and not skewed). Merging probably helps with performance too.

In case anyone wants to take a closer look at similar mesh alignment seams in-game, the map in the screenshots is DM-Nitro.

Saturday, 23 July 2016

My Favourite UT4 Maps

Unreal Tournament 4 already has a large number of custom maps, remakes and ports out there. These are my top 3 personal favourite maps to play in UT4.

DM-Broken - by ConradJustin

"This is my shield, I dare you to try taking it from me!"

This one really stands out from the crowd. It looks good, plays great and seems to be made up entirely of custom content made specifically for the map.

Broken is a fairly large map with plenty of vertical action. Lots of ramps, stairs, elevated platforms and walkways, a few lifts and several well placed jumpads. It has good connectivity and sightlines without being over connected or having overpowered sniper camping spots. There are plenty of chokepoints and strong defensive positions for more strategic play. Most of the combat happens at short to medium range, there aren't a whole lot of really long sight lines.

Overall weapon and item placement is very good too though one thing I don't care for is the stream with a keg and a bunch of health vials all lined up. IMO that's just too much over health to have all in the same area.

I really like the shield area. The upper platforms around the shield are a great camping spot but not so great nobody can knock you out of there (especially not if they gang up on you). It works out to a sort of king of the hill style game. Sure it's a strong spot to hold, but if you're up there you aren't getting much overhealth, armor and probably not the jumpboots or anything else either, all of which make a successful counterattack more likely.

I really like playing this map King of the Hill style, controlling the shield, the area around / below it and daring anyone to try taking it away from me.

Not wishing to give anything away, but let's just say it's well worth your time to search every nook and cranny looking for hidden secrets.

DM-Nitro - by me

You might think I'm biased with this choice since I remade Nitro for UT4. It's not that I like Nitro so much because I remade it, just the opposite, I remade Nitro because I like it so much.

The map is appropriately scaled up (2.5 times larger), with proper UT4 lighting, materials and reflection environment. It's mostly meshed now, though a lot of the large floor, wall and ceiling surfaces are still BSP.

Layout is the same as in the original, item placement is mostly the same with a few changes and additions to address the problem with one side of the map being so much stronger than the other.

Whether the game is DM, TDM or Duel, it plays great in UT4.

DM-CurseII - by MetalFist

An old favourite from UT99 this one's more of a port than a remake, appropriately scaled up by the same 2.5x factor but still using the old UT99 textures.

Other than not having a Ripper, layout and item placement are identical to the original. Gameplay is about as good as it gets, it plays great!

Curse was never one of my favourites in UT99 but it certainly is now in UT4. I love playing it now, maybe that's because I'm more of a shock ho now than I used to be and the top of Curse is a great place for shock ho'ing.

Sunday, 3 July 2016

Lightmass Swarm Farm Growing

I picked up a used Dell Optiplex 790 SFF for $140 CDN to add to my UT4 Lightmass farm today.

The machine came with an i5 2400, 4G of RAM (2x2G) and a 360G HDD. I added a another set of 2x2G RAM I had lying around to give it 8G total RAM. That's probably overkill, 1G per core or 4G in total should be plenty but since I already had the RAM anyhow, better safe than sorry.

That gives me a 3rd machine with a little over 6100  passmark performance test cpu benchmark score to add to 6300 from my i5 4690 main computer (the main machine actually scores 8400 on all 4 cores but only uses 3 cores to build lighting, so 8400 x 0.75 = 6300) and 4600 from a 3rd machine with a Phenom II B93.

That gives me very nearly 2.7 times as much multi-threaded processing power to build lighting with as with just the main machine alone (again only 6300 since it only uses 3 cores).

6300 + 6100 + +4600 = 17000
17000 / 6300 = 2.698...

IOW, my longest build time (DM-Eliminator with production lighting quality) is about 2.5 hours with just the main computer (6300 cpu benchmark score) working on it and 1.5 hours with both the main computer and the Phenom II working on it (6300 + 4600 = 10900 cpu benchmark score). With the new machine added to the farm bumping the total available processing power to 17000 cpu marks it'll take just under 1 hour (55.6 minutes according to the math) on production lighting quality.

Edit: Rebuilding Eliminator with all 3 machines came out to 57.5 minutes, within 2 minutes of what the math above predicted.

Now just imagine what oh say four OC'd 5960X machines with ~80k cpu benchmarks would do to build times. =)

Saturday, 11 June 2016

DM-Nitro Progress

I've been working on DM-Nitro lately.

I did quite a bit of meshing, another lighting pass and some optimization too (cut the filesize nearly in half).

I'm mostly using stock assets in Nitro to keep things simple, clean and readable for visual clarity and performance. I'm making heavy use of skewmesh splines and merged meshes too.