• Hey, guest user. Hope you're enjoying NeoGAF! Have you considered registering for an account? Come join us and add your take to the daily discourse.

Halo 4 Forge |OT| Crafting a better tomorrow; one Brace, Large at a time

Thanks for the invite, GP!

We played a hilarious round of Big Team Slayer on Rocksteady (my Avalanche remake) and it seemed pretty fun. There are some obvious caveats to the map, like the top section that I've added a kill wall to (should I remove this, and let the banshee fly over the top?) and the especially steep ramps up to middle tunnel (a product of scrunching the level into Ravine, I think), but overall I think it was a good time.

Any impressions? Comments? Thoughts?

I remember someone mentioning some peculiar respawn points, so I'll have to do a sweep and make sure there's nothing too out of place. Overall, I think the spawns are a lot better on CTF, but it seemed silly for the debut of the map to require objective, especially with all of the lag we had. It'd be nice to try the map with more regionally similar people to get a better idea of how it plays, as I don't think it was the map itself causing the lag.
 

Computron

Member
just gonna move this up since it is right up against the end of this last page.


---------------------------------


So, about that lighting write-up...

Lighting is definitely important, and I hope I can adequately explain what I mean to show, despite all the problems and the not so stellar examples. I got no capture card, no fileshare access, and these examples take a long time to make and these are not even the best ones by a long shot. (I will probably do more eventually) To make these .GIFs, I took a screenshot, moved a forge pieces just a bit, switch to player mode and waited for it to generate lighting, then I took a screen shot again and moved the pieces again, generated lighting and repeated this for every frame that I later comped in AE into .GIFs.
--

The ambient/indirect lighting that is generated/baked onto your pieces helps your players with navigating your map by provides depth cues and visually separating the walls from the floors and ceilings as well as aiding in the process of judging distances and player positions. It is also useful for subtly guiding your players where you want them to go:

Click on Image to Expand to full size

iGrAYrlri7b2I.gif
As you can see, the direct sunlight and the dominant shadows are not the only aspects of lighting that are important, the walls and floor also receive a lot of 'indirect light' from the (non-forge piece) environment around them such as the sky.
(And, oddly enough, pretty much everything else in sight, even stuff that does not make sense like grass that is in entirely in shadow, but more on that later...)
So, when I say ambient/indirect shadows, I am referring to the colored gradients of shadow, or the shading, that is seen inside the dominant shadows that are cast onto your forge pieces. Most importantly, the ambient/indirect lighting is what gives your forge pieces most of their color, especially on Ravine Forge variants. On Erosion variants, there is very little direct lighting so indirect lighting is also crucial. On the other hand, on Impact variants, there is very little effect from the indirect lighting, which is a shame.

So when your forge pieces get packed really close together and start to occlude the sky and surrounding environment around them, the ambient lighting is not able to reach into the crevices, folds and various levels of multi-storied buildings so the ambient lighting starts to become very flat and monotonous.

You can start to see this in the .GIF above, where the arrow is pointing on the floor when the building moves close to the raised floor on the left side of the .GIF:

wObYq.png


The ambient lighting has a very short reach because Halo 4’s ambient lighting system is a type of single pass radiosity, it does not include any form of diffuse interreflectivity between forge pieces where the direct sunlight can bounce off the floors and add-to/fill the room, and it can only really be seen subtracting ambient light inside the dominant shadows. So if you tightly enclosed spaces, your ambient lighting will become very flat and difficult to navigate very quickly. When most of your map is like this, that is when you run into the problems of completely monotonous, ugly, solid dark-brown lighting, such as I mentioned at the beginning. I see this way too often on many people’s maps, especially the ones I have seen on YouTube and ForgeHub. It can be difficult problem to overcome in tight interior maps, but in most cases all you need to do is make a window looking outside or open-up/raise the ceilings, because, even if it does not let in the sun, it will still add ambient/indirect light.

---

Also, the games engine does not spend a lot of time generating the lighting/shadows when you switch from monitor to player mode, so they do not get a lot of samples, which makes the indirect lighting look all segmented and polygonal rather than the smoother gradients you would get from more samples.

The lighting info is not baked into the actual map file that people save to their hard drives and put on their fileshares, but rather it is regenerated every time you load the map. This is almost certainly done to save space by reducing the Forge variant's file size. So, just as a speculative aside, this seems to me like they might have the opportunity/possibility to make the lighting better quality by simply cranking up a few values behind the scenes, maybe in a patch. Probably not, but its just a thought. These parameters would make the lighting generation time a bit longer, increasing the number of samples it would generate, but increasing the quality significantly. This could perhaps be limited to only the times when you are trying to play your final map outside of active forging. Every MM match has a ten-second countdown timer before the game starts anyway, and the lighting generation only uses a small portion of this time. So, hey, why not use that extra time to get some more samples for some better lighting? Maybe there are other reasons that preclude this from being a possiblity, but I don't know.

---

For the foundation remake, I set up the bases so that the warmer-colored parts of my map would be in the direct sunlight. This means the red and yellow team's bases would be very bright and easy to see, placed in the direct sunlight. On the other side of the map with the colder colors, it started out almost the same way. It was originally pretty bright because I was remaking foundation, which is an open top/sky map, and the way the sun is angled on Ravine let in a lot of sunlight:

Direct sunlight over some parts of the blue and green teams half:

Click on Image to Expand to full size

1Q1cB.jpg

I decided it would be better to cover that half of the map in shadow so that the colder color team's, blue and green’s bases would match with their lighting and you could naturally orient yourself. I could not move my entire map to a place where this would naturally occur from the terrain, since this would take a long time and it is quite hard to find a spot that will do it correctly and have enough room for the entire map. The shadowed, outside half of ravine is quite glitchy with its shadows. For example, the large forerunner structure has some strange problems where it will incorrectly bake its shadow unto your forge piece. Using the map geometry for lighting and shadowing also has some other problems that I will address later, particularly with indirect light, which they seem to handle with some cheap approximations. (Like the example I posted earlier where having a completely closed off box with just a little sliver of grass visibly sticking up through the ground plane of the box, somehow shining in a ton of green ambient light. IDK what is the deal with that, I guess the grass is radioactive?)

I wanted some more control over the lighting in my map so I ended up using some coliseum walls to block off the light from the sun and cast a shadow on the colder side of my map. I set up the wall piece to cast a shadow only over one-half of the map and stop in the middle when it starts to cross over into Read and Yellow team's base:
Click on Image to Expand to full size
iHHU1TfwrzAQ8.jpg
The shadow seen over the blue/green team's side is cast by the coliseum wall piece that I placed far behind the camera

By controlling the distance of the coliseum wall away from the spot that it is receiving the shadow, I can control the softness and gradient of the ambient/indirect lighting inside of the shadow, while maintaining control of the shadow itself. The shadow stays in the same place and covers the same half/portion of the map, as intended, but the ambient/indirect lighting inside the shadow can be adjusted to be easily readable and soft enough to still provide depth cues and separation without clipping into flat dark grey-brown:

Click on Image to Expand to full size

iUaUF2jNLfzvI.gif
The softer and lighter the lighting gets on the back walls and around the turret overhang section, the farther away the coliseum wall was at that point in time in the .GIF.

The only thing changing in the picture is how far above the coliseum wall (not shown) is from the top of the wall at the back of the map.

The effect here is pretty subtle, but you can use this idea/technique to get your ambient/indirect light to have a farther/deeper reach. In a way, it can also be sort of like targeted-adjustments/color-correction in photoshop. Just be aware that the ambient/indirect lighting tends to clip into flat, dark-brown very quickly.

Anyway,I hope I explained that adequately.
 

op_ivy

Fallen Xbot (cannot continue gaining levels in this class)
Planning to start customs in 10 minutes! Anybody else have maps to test? I downloaded GP and op_ivy's.

did you guys get to try out resonance before i joined? i had to quit, the BTB lag was horrific.
 
any invites i'm down
GT - Archpath1

Not sure if the lobby is still going on. I had to back out a little bit ago after tons of games. :) Had a blast. Digital has got his feedback, I've given Crow my feedback, and all that is left is Ivy. :) Also Raide we didn't get a chance to play your map huh? :(


did you guys get to try out resonance before i joined? i had to quit, the BTB lag was horrific.

Yes we did get the chance to play it. Many people felt fond of the center area which was cool. The outer corners felt like a labyrinth of hallways with lots of railings and felt like we had to jump everywhere to get to places. Also the Tower, 2 Stories that you have on the corners don't have a railing on the lower ramps and resulted in a couple of people falling off the map. Had a blast on it still though. Fooled a ton of people with Hologram.


Thanks for the invite, GP!

We played a hilarious round of Big Team Slayer on Rocksteady (my Avalanche remake) and it seemed pretty fun. There are some obvious caveats to the map, like the top section that I've added a kill wall to (should I remove this, and let the banshee fly over the top?) and the especially steep ramps up to middle tunnel (a product of scrunching the level into Ravine, I think), but overall I think it was a good time.

Any impressions? Comments? Thoughts?

I remember someone mentioning some peculiar respawn points, so I'll have to do a sweep and make sure there's nothing too out of place. Overall, I think the spawns are a lot better on CTF, but it seemed silly for the debut of the map to require objective, especially with all of the lag we had. It'd be nice to try the map with more regionally similar people to get a better idea of how it plays, as I don't think it was the map itself causing the lag.

Glad to have you there bro. Too bad you couldn't stick around to play a game on Think Again. :p You got my feedback during the game on Rocksteady. ;) Had a blast, minus the lag.




Any feedback on Think Again for those of you that played it?
 

enzo_gt

tagged by Blackace
Computron, your foundation remake is amazing. Rooms still need work (I think?) but it's so beautiful already.
 

Raide

Member
Not sure if the lobby is still going on. I had to back out a little bit ago after tons of games. :) Had a blast. Digital has got his feedback, I've given Crow my feedback, and all that is left is Ivy. :) Also Raide we didn't get a chance to play your map huh? :(

Its all good! I dunno if my MIC was working ok. I could hear everyone but not sure if I was coming through. My map is a 4v4. Any more than that and it would be uber hectic. I do have a BTB (Possibly) map I was working on, so I will have to get the spawns ready for it. :D
 

op_ivy

Fallen Xbot (cannot continue gaining levels in this class)
Yes we did get the chance to play it. Many people felt fond of the center area which was cool. The outer corners felt like a labyrinth of hallways with lots of railings and felt like we had to jump everywhere to get to places. Also the Tower, 2 Stories that you have on the corners don't have a railing on the lower ramps and resulted in a couple of people falling off the map. Had a blast on it still though. Fooled a ton of people with Hologram.

nice. i had only got to dick around 1v1 with my nephew before sharing it. thought the mix of long and short sight lines coupled with the smaller size would be a nice change of pace from the larger 4 maps.

having those side towers was to help keep some areas more focused on CQC, but yeah, their railing was an issue. should i scrap the towers for something else, or was it just learning the new map?

We did. You can find a film of it in my file share. Let me know when you pull it down

i'll get it tonight and send you a pm or msg on live. thanks guys.
 
nice. i had only got to dick around 1v1 with my nephew before sharing it. thought the mix of long and short sight lines coupled with the smaller size would be a nice change of pace from the larger 4 maps.

having those side towers was to help keep some areas more focused on CQC, but yeah, their railing was an issue. should i scrap the towers for something else, or was it just learning the new map?



i'll get it tonight and send you a pm or msg on live. thanks guys.

I highly suggest creating your own structure. Pre-built structures are just that... pre-built for a particular purpose for their respective map. You can never guarantee that it will work perfectly for what you want and it also pulls away from originality of a map. I run by the guideline of never using pre-builts as they were intended. To start I'd say make your own version of the tower with individual pieces and then adjust as need be.

As for just learning the map, if users have trouble on their first run it is a safe bet to address that issue as soon as possible. That first impression is the most important thing you have for maps. If players have any sort of frustration it will ruin that. Plays off the philosophy that it only takes one bad thing to ruin a ton of great things. ;) So if there is a chance to make it better then take it. ;)
 

FyreWulff

Member
For the record, the following maps don't support Dominion objects of any type:

Complex
Adrift
Haven
Solace
Abandon

I wonder if the non-Forgemaps with Dominion objects were the ones Certain Affinity made, since Certain Affinity made the Dominion gametype:

Ragnarok
Exile
Meltdown
Vortex
Longbow

We already know 100% for sure CA made all the Forge maps.
 
For the record, the following maps don't support Dominion objects of any type:

Complex
Adrift
Haven
Solace
Abandon

I wonder if the non-Forgemaps with Dominion objects were the ones Certain Affinity made, since Certain Affinity made the Dominion gametype:

Ragnarok
Exile
Meltdown
Vortex
Longbow

We already know 100% for sure CA made all the Forge maps.

That is an interesting point. We know that Kynan made both Complex and Haven. Abandon was talked about being in office as well. Not sure on Solace and Adrift. It wasn't specific on whether or not Meltdown was done in house, but there was talk of having an Infinity loop map on the todo list. So who knows. Interesting deduction for sure. Also Ragnarok is the only one on that list that isn't in Dominion if I am correct. Could possibly be that those were the only maps they expected Dominion to be played on.
 

FyreWulff

Member
It'd help explain why such a big map like Complex doesn't support Dominion. And why Ragnarok would support it even if it wouldn't work that great.
 

Striker

Member
It'd help explain why such a big map like Complex doesn't support Dominion. And why Ragnarok would support it even if it wouldn't work that great.
It still wouldn't explain why an asymmetric map like Complex is getting used in 2-flag too, but that's another thread.
 

FyreWulff

Member
It still wouldn't explain why an asymmetric map like Complex is getting used in 2-flag too, but that's another thread.

Neutral Bomb on Ghost Town.jpg

Anyway, I've got a couple more Flood variants of developer maps on my file share. "Decomp" is my Flood variant of Complex, while "Facemelter" is my infection variant of Meltdown.


Decomp:

* Open side of the map is now blocked off
* Extra clutter and flow changes
* Exploitable spots softkilled off
* Mancannons to make the high point more assailable by Zombies
* Based off the latest MM version of Complex, so you can't jump into the geometry


Facemelter:

* Meltdown cut down to about 60% of the original map.
* Flood appropriate scenery, cover, and clutter
* Turrets!
 

FyreWulff

Member
Unfortunately they need play intervention to actually spawn.

So some further tests with Pete and Hyokin:

- Turrets in Infection cannot be made to only shoot at zombies, they'll shoot at both player types. However, if the turret is set to Attacker (Zombie) or Defender (Humans), only that player type can initiate the build.

- Dominion shield doors revert to neutral doors when in Infection.



Useless Forge quirk:

If you add a few objects to Meltdown, it'll pause for a moment like it's about to generate light shadows ala the Forge maps but won't end up generating any.
 

Raide

Member
Something that bugs me the most in Forge. When I copy an item, the camera gets closer and closer. Why the hell did they make it do this? lol
 

Raide

Member
Done some work on my 4v4 map and my BTB map. Had to add a bit more to the BTB map to make sure it won't be too cluttered. Still trying to work out the Fileshare problem I have, then I can upload it.
 

FyreWulff

Member
Petetheduck discovered that Halo 4 blocks anything from spawning if a player is standing in it's spawn space. Including structure objects. So when he went to try and remake the Musical Chairs gametype, the next 'level' wouldn't spawn since the players were inside where the object would exist.

In Reach, things would just spawn into you and kill you if you were in the way. So if you need something that delays a spawn to work or even things like the Dominion shield doors won't come back if you're standing where they'll spawn.

Note that when you're forging.
 

Computron

Member
Cool, so now we can have a dead mans switch. Also, couldn't your partially replicate what your looking for with a movable object like a crate with its spawn point floating in the air? I know it's not the same thing, especially if you need something the size of a coliwall but still.
 

FyreWulff

Member
What 343 Should Do To Improve Community Content Generation Going Forward


Everyone’s said their piece here and elsewhere about the downfalls and improvements of Halo 4’s Forge. MrBig’s gigantic post addressed a lot of features and tools that could be used within the current Forge system, but I’ve put some thought to it and want to make the following post/suggestion for the next step in controller-based community content creation. Of course, letting people just use the existing Halo tools will always be the best option, but within the confines of the console and controllers, there’s still plenty of room for improvement and flexibility.

With that in mind, I propose splitting map creation into two separate tools, Forge and Anvil. Diagram, courtesy of Devolution:


MnOf5.png


Forge as a level geometry editor (Halo 3 Foundry and Sandbox, Reach/4 and the STRUCTURE category) was useful for the time, but the limitations are starting to impact actual map development. What should be done is to return Forge to what it started out as in Halo 3: an easy to use object editor that players use to modify objects on existing maps. Anvil (stealing the Marathon tool name) will become the actual geometry/world editor to give map makers more power in creating their world, while making it easy for beginners to easily alter developer maps and community maps.

Anvil would be a single-player only experience. No networking or multiplayer real time tank dropping, let’s leave that over in Forge. Talk to the guys at Ubisoft about their editor, but Anvil will start as a blank slate with basic heightmap editing:

omgfarcry2.jpg



While having Terrain-style objects in 4 would be nice (grassy hills, more useful rocks, trees and such), the problem still exists that the engine treats them as added objects that it has to aggressively drop detail on / dither in and out at range. Anyone who played 4 player split in Reach on Hemorrhage remembers everything popping in and out of view. Imagine that with actual map terrain. Hell no.

This is where the base map geometry/land would be made. Editors would define the water level, the mountains and valleys, trees, and structure objects like buildings and so on. While being able to currently rip apart a map with your buddies in Forge in real time is fun, you rarely need to or want to change your map’s base geometry after you finish it.

This mode would also act like Far Cry in that you’d be able to paint grass, rocks, cliffs, and add Cave objects to make negative space out of the height map (something the Far Cry Editor supported on the original Xbox in 2004). Then you’d set the time of day, the angle the sun is at, and select a skybox, which would determine the overall ambient light on the map. Day/Noon/night/dusk would the the main options, although the FC editor and others let you fine tune what angle the sun is at.

Finally, the forger would be able to add much more lights inside of the map. Currently in 3/Reach/4 Forge, the lights are objects that are placed as objects and barely light anything. In 4, they aren’t considered in the lighting pass done on objects. And they’re not rendered in splitscreen. Waste of budget, performance, and not even consistent for appearances. Anvil lights would be used to light up the insides of buildings, caves, add athmospheric lighting and many other things you can do with lights. They would be baked into the shadowmap when the engine goes to generate it, and would not be rendered as glowing balls in gameplay. Since the engine is no longer generating light in real-time, we can have more of these points of lights in Anvil and you’d see the lighting effect in splitscreen.

fPmK8.png


You’d also set your ‘hardcoded’ world soft/hard kill zones in Anvil. For beginners, you’d add some random terrain/foliage generation and presets in the menu to start with. You could even make some presets like “Blood Gulch”, “Sidewinder”, “Containment” etc to get people started on some basic map types. After finishing your map and finalizing it, the game would bake the lighting and do some basic geometry welding / optimization. Your new map would be ready for the Forge process.

Forge would return to being a basic editor for spawns, gametype-specific objects and soft / hard kill zones and placing of any objects that move dynamically. Some structure-type objects are still available but are still the basic FW pieces for closing off / opening parts of your map for game modes and really basic map structure editing. This is where you’d do Gametype rigging, and if 343 allowed it, rigging of AI placements, movement maps and such for Firefight. Once you’d be finished, you could go back to Anvil and select your Forge variant as the “Built In” variant of your base geometry.

So say I was remaking Blood Gulch:

* Make Blood Gulch in Anvil. Save.

* Make Blood Gulch’s objects, spawns, etc in Forge. Save.

* Open Blood Gulch in Anvil, Assign Default Forge Variant -> Blood Gulch.


Anvil variants would show up in the top level menu alongside developer maps, with their Forge variants (with at least one built-in required) as a sub selection, just like in 3/Reach/4. So you’d go down the menu and see


Anchor 9 > Anchor 9*
Vortex > Vortex *
The Pit > The Pit *
Pitato
Charnel Pit
Cust Map > Cust Map *
Cust Zombies

Next post: Packages
 

NOKYARD

Member
Petetheduck discovered that Halo 4 blocks anything from spawning if a player is standing in it's spawn space. Including structure objects. So when he went to try and remake the Musical Chairs gametype, the next 'level' wouldn't spawn since the players were inside where the object would exist.

In Reach, things would just spawn into you and kill you if you were in the way. So if you need something that delays a spawn to work or even things like the Dominion shield doors won't come back if you're standing where they'll spawn.

Note that when you're forging.
I should have mentioned that little fact sooner since it was the very first thing noticed within minutes of getting the game.

In live broadcasting Reach games we relied on the forced spawning of items to push the cameraman into position inside a Coliseum Wall by spawning another wall behind him (pic is the camera position inside the wall). Once inside he had an unobstructed view of the playing field while also being invulnerable. We even had a complicated system to bifurcate the cameraman, commentator, and players using the 2 different player models. Players (Human) spawned in the game, and camera (Elite) spawned inside the wall. From the player's point of view it was a regular game with regular spawning. The courts could even be used as normal courts with default variants as long as no one was a dino.

All of that is now impossible because of this new item spawning system. There seems to be no way of replicating any of these in Halo 4 so we had to go back to our Halo 3 method of spawn room and having players run into teleporters which placed the players onto the field. Now, the cameraman is stuffed inside the corner cylinder of a Triple Room which seems to be the only forge item where you can place a receiver node. Even the closed Containers we used as camera boxes in Halo 3 are no longer useful since they are now textured on the inside.

If anyone wants to know what the hell i'm talking about, download Primordial LIVE ( 4v4 + 1 Camera) from my file share and walk backwards at spawn in a team game variant like Team Slayer or Grifball.
 

FyreWulff

Member
Packages

Along with upgrades in the editing, we need upgrades in the distribution methods. While we had Bungie.net Web fileshare links and File Sets, and hopefully some Halo 4 web shares eventually, they're still did not solve all the problems. If any of you have ever released a minigame + map combo, you've noticed that you always seem to have roughly half the downloads for the gametype as you do for the map. So there's a bunch of people playing your map out there.. somehow.. in a completely useless state. I had a Gametype + Map combo on Bungie Favorites at one point called "Flytrap" and included "Sandtrap Fly" next to it in the listings when I submitted it to Urk, and still got tons of "I downloaded Sandtrap Fly, now how do I start flying". Facepalm.


The solution is something better than File Sets: Packages. What are Packages, you might ask?


A Package is essentially an end-user method of making Map Packs like 343/Bungie can when they sell DLC. It resides as it's own object in your File Share as a single slot.

A Package can be one of three types:


- A Gametype + Map Package

- A Map Pack Package

- A Collection



A Gametype + Map Package is slightly different but uses the same concept. You create a Gametype from the Gametype editing screen. You can then choose to "Attach Map.." and can select your Forge variant (and the Anvil variant if it's attached to one), which will automatically create a Gametype Package for you that you can upload to your share. Someone can then download it.

So where in Halo you have the options:

Gametype: <type>
Map : <map>
Start Game

You'll have

Gametype: <type>
Map : <map>
Package: <package>
Start Game

Selecting the Package will automatically bring in the map and the gametype.


A Map Pack Package is simple. You create the package in-game. You name it, add a description and add a thumbnail. You then select maps from your local storage, hit Y, then "Add to Package". Once your package is finished, you'll upload it to your fileshare. This Package will now have a unique server ID. A Map Package is different in that the Package is just for maps, and these maps are always "together". If you delete one map from the package, you delete the entire package (with a notice that you're doing so). Remember the ID number I mentioned earlier? Packaged maps can also be updated; each time you change your Map Package and upload it again, everyone's box with that package polls Waypoint for updates, and if they see it, they'll download the package, delete the old version, and put the new ones in it's place. If people want to keep a map in a Package permanently, they can Save As and make a copy.


So not only can we make proper minigames, you can also make proper map packs, and also ship updates to maps via the packaging system. Maps and gametypes can still be individually uploaded, of course. There's already a form of bundling going on currently since Halo 3; if you play a multiplayer film, the film contains the map variant and gametype and will populate your Recent History for both with the map/variant from the film. So it's not that far of a stretch for the engine.


A Collection is closer to the File Sets of Bungie.net but can contain any file type. Initiating this essentially just auto-queues a collection of files as a download. So someone could collect a bunch of their favorite maps and gametypes together and point people at it. The improvement over Bungie.net is that you wouldn't have to start each file individually, and you could download it from in-game.
 
Anvil would be a single-player only experience. No networking or multiplayer real time tank dropping, let&#8217;s leave that over in Forge.

IMHO this is the one thing that kills the idea for me. Over the past 5 years, the single best thing that has improved map design in the community is collaborative ideation with a group. Being able to adjust the structure right there on the spot to encourage others to come up with ideas has been a key thing that makes Forge one of the better tools out there for learning. More control over the terrain, structure, and lighting is NOT worth the sacrifice of that collaboration. I've seen hundreds of new forgers grow quickly because of this sort of collaborative forging. I've seen forger's block solved many times over by having a small group get together and start tossing around ideas by throwing down blocks. Placing one block could spawn a dozen ideas for the next block and so on and so forth.

It is one thing to just say "what if you tried this" and another thing to actually do it in front of their eyes. The value of that feature is not worth removing for anything, IMO.



This I can get behind, but I think it could be any combination of maps and gametypes, because several established gametypes have also had variants tied to them as well.
 

FyreWulff

Member
IMHO this is the one thing that kills the idea for me. Over the past 5 years, the single best thing that has improved map design in the community is collaborative ideation with a group. Being able to adjust the structure right there on the spot to encourage others to come up with ideas has been a key thing that makes Forge one of the better tools out there for learning. More control over the terrain, structure, and lighting is NOT worth the sacrifice of that collaboration. I've seen hundreds of new forgers grow quickly because of this sort of collaborative forging. I've seen forger's block solved many times over by having a small group get together and start tossing around ideas by throwing down blocks. Placing one block could spawn a dozen ideas for the next block and so on and so forth.

It is one thing to just say "what if you tried this" and another thing to actually do it in front of their eyes. The value of that feature is not worth removing for anything, IMO.

Networking would stay in the Forge side. Structure pieces would remain on the Forge side as well - the ones in Anvil would be more feature-rich and complex.

We're just going to have to disagree though, I'd rather have a more powerful toolset over a networked map editor session. The only time I've built a map as a Forge session was me and Pete's Relay premake.

This I can get behind, but I think it could be any combination of maps and gametypes, because several established gametypes have also had variants tied to them as well.

Adding a third package type.. I think that will cover all bases :)
 
Networking would stay in the Forge side. Structure pieces would remain on the Forge side as well - the ones in Anvil would be more feature-rich and complex.

We're just going to have to disagree though, I'd rather have a more powerful toolset over a networked map editor session. The only time I've built a map as a Forge session was me and Pete's Relay premake.



Adding a third package type.. I think that will cover all bases :)

If the networked Forge side keeps the structure set then I'm down for it. I wasn't just referring to forging as a group though, but getting direct feedback from peers. After a lot of playtests, several people will go to forge and offer suggestions while in Forge by throwing down their ideas for everyone to see and piggy back off of. Or when someone gets forger's block they will invite friends to have a look and people will through down their ideas. That has been a crucial feature that makes Forge for me.

I approve of said third package. ;)
 
Great ideas Fyrewulff. I've had a similar idea for a while now even down to calling it anvil. I really hope this is how the next game is. Plus anvil is an excellent name for the editor.
 

Raide

Member
I am all for anything that helps the Forge community expand. I just hope they put enough thought into the actual tools and not just the back-end fancy bits. I will be really annoyed if Halo 5 drops and its just another Reach Forge. I really want to see them push the tech on and make a standout Map Editor for consoles.
 

Computron

Member
What Firewulff means to say is, release the H4EK with full on 3ds max and photoshop exporter plugins and let us write our own shaders and game scripts.

:p

Screw the controller. I was thinking of making a sketchup exporter/converter with the forge map variant file structure, but it seems like that would be limited in its utility because they appear to have server sided checks for any custom content, so you can't really mod anything outside of JTags.
 

FyreWulff

Member
What Firewulff means to say is, release the H4EK with full on 3ds max and photoshop exporter plugins and let us write our own shaders and game scripts.

:p

Screw the controller. I was thinking of making a sketchup exporter/converter with the forge map variant file structure, but it seems like that would be limited in its utility because they appear to have server sided checks for any custom content, so you can't really mod anything outside of JTags.

At one point I was conisdering making a program where you'd make a Forge variant in a program, then it'd output buttons to a 360 controller and build it in Forge. Testing showed that the pieces never moved quite the same distance and the coasting bug in Reach and 4 prevents using inputs on the coordinate system. Mad scientist shit, yo.


Also, pete showed me in a custom that Flood can thrust right through one way shields and Dominion shields. Note this when you're Forging (good thing I already didn't trust them and had kill volumes behind them when bordering my maps)
 
If the networked Forge side keeps the structure set then I'm down for it. I wasn't just referring to forging as a group though, but getting direct feedback from peers. After a lot of playtests, several people will go to forge and offer suggestions while in Forge by throwing down their ideas for everyone to see and piggy back off of. Or when someone gets forger's block they will invite friends to have a look and people will through down their ideas. That has been a crucial feature that makes Forge for me.

I approve of said third package. ;)

Oh. My. God.

Imagine Forge with Version Control!

Submit a pull request, clone the file, make changes, push them to the repo, resolve any conflicts/diffs.

*nerdgasm*

Version Control all the things!
 

Raide

Member
Ok, finally got the Fileshare Upload working for my maps.

I currently have 4 in there. 2 are basic layout that I am messing with, might be nothing in the end. The other 2 are further along and looking good but I am trying to get some feedback.

Tower and Hole and the real WIP maps with basic spawns and basic layout.

Verith (4v4) (Jonestown for anyone that saw/died on it) and SpaceBase (BTB)

For Verith, I am wondering if it is ok for 4v4 or if I should expand it to BTB somehow. It has a DOTA feel/look because of the lanes and distinct colouring. At least its not grey!

SpaceBase is much bigger with the layout done but still a few empty rooms until I can figure out what to put in them. Not sure if its big enough for BTB thought.

Any thoughts would be welcome. I need to get back into Forge and start making some better maps.
 

Computron

Member
At one point I was considering making a program where you'd make a Forge variant in a program, then it'd output buttons to a 360 controller and build it in Forge. Testing showed that the pieces never moved quite the same distance and the coasting bug in Reach and 4 prevents using inputs on the coordinate system. Mad scientist shit, yo.

No you didn't, really? Why would you first think to approach the problem through emulating a controller though? That seems like a crazy complex way of doing it.

But you say you tested this, right? So you have a system in place to emulate a 360 controller? You should make my theater 1080p recorder a reality then:

[snip]

Wish I had a good capture card. But, It looks great in 1080p with the extra theatre mode enhancements!

Now that I think about it, you could set up an Arduino to record halo gameplay footage in 1080p with perfect anti aliasing and theater-mode-extras like longer/higher LOD distances, better filtering and potentially infinitely higher frame rates for procedural effects like Particle/water/plant-animations effects (though, sadly, not player animations since those don't seem to get interpolated). Maybe you could even capture it all in stereo 3d through a slight offset. All you would have to do is get the arduino to emulate the Xbox controller with some macros to automate all the button presses and capture everything frame by frame.

Man, that's gotta be one of the craziest ideas for capturing 6 yr old console game footage. What did Bungie call their video capture service? I think the internal code name was the Hydra system? I need a name for this one for when I launch my own service in direct competition. Lol, charge $20 for a minute long true 1080p clip of Halo .4 :p

[snip]
 

FyreWulff

Member
No you didn't, really? Why would you first think to approach the problem through emulating a controller though? That seems like a crazy complex way of doing it.

But you say you tested this, right? So you have a system in place to emulate a 360 controller? You should make my theater 1080p recorder a reality then:

The hacked up controller I was using broke.. use a wired controller and you can hijack the shit out of it.

The general idea (I figured) was that I could always start at a known point, so all I had to do was make the actions repeatable with the same result. Then it'd go through the menu, spawn an object, move it, return to the origin point, spawn an object, move it.. issue in Reach is there's the coasting bug. What I'm talking about is if you move an object via the coordinate system long enough, say you're moving it along the X + axis, if you try to go in the other direction on the X - axis it'll still keep moving in the X + axis for a bit for some reason. This bug is still in Halo 4. I really wish the coordinates were absolute and you could just punch in the numbers and the object would be teleported to it's new origin. Then you could just make a builder program that spits out the correct coordinates and tilts.

My favorite hack I still did was playing Halo 2 on Live via a GameCube controller.. :p

Could be something to consider for a future project though. Sounds relatively simple enough, since films are linear. It'd help if we had a true "frame by frame" button. Emulate button (currently, a quick trigger tap I guess), press A to capture screenshot, wait x seconds, capture buffer, discard screenshot, advance to next frame.
 

Computron

Member
Sure, do it if you want! You just need to emulate a very light trigger pull and set the rest of the buttons up on timers.
I already got a system set up for sound capture as everything is already routed through my PC.
All I got to do is load up Audition and record a normal play through of the film and then sync the sound up in After Effects. It would be cool to make that part automatic as well.
Also, what is the max amount of screenshots Halo 4 can write to your HDD or Flashdrive?
I think its still lower than 100, so that would have to be dumped onto the PC approximately every 3 seconds of in-game footage... more stuff to automate, cool.


Also, I got to see this gamecube controller you speak of, got any pics, or better yet, video?
 

FyreWulff

Member
Nope that was well before I could afford having a digital camera. It was a basic wiring job of a battle-worn GC controller into an Xbox controller with shitty sticks. I had shitty <512MB ram computers until late 2008-ish. This was the last frankenputer before my fortunes changed and got a good one from a contract job:

4BnM3.jpg


I was always basically scavenging parts off other dead computers just to stay on the internet and in contact with people and other connections. Honestly if I hadn't, I'd probably be homeless now. That's a bit deep for a Forge thread though :p

But as pointed out above, you could probably do that real nicely today with Arduino and such. It could output the controller buttons AND send messages to the computer to control After Effects or anything other program.
 

Karl2177

Member
I'm gonna need a refresher. Why are certain ordnance drops labelled with Attacker or Defender? I was under the impression that they were supposed to be neutral.
 

External Memory

Neo Member
^

If only Valve/Source hadn't already taken the level editor name "Hammer".

Hammer and Anvil would have been awesome.

Not sure if Foundry is copyrighted for a level designer, but Foundry, with subcategories Core (for your Anvil) and Caster (for your Forge) and Typographer (for gametype lol not fonts but there's another metaphor) for custom game editing would be pretty tight for Destiny (I'm looking more forward to what Bungie has to offer now honestly 343's honeymoon's over).

The other idea of having updateable packages linked to maps or map/gametype combos could still work, I'd call them Shipments since it implies a more active, updating thing.

So every map would be made of a Core and at least one Cast, while Shipments would consist of a Core, (at least 1) Cast(s) and, optionally, one or more Types. The system would ideally have as close as possible parity to how devs update maps in MM, so they can autoupdate "regulation" maps you can access from your own library without swimming through MM to get the latest, and you could optionally allow for old shipments to not be overwritten, you just would be notified if you run out of storage.

The names would also help people recognize/specialize into their favorite roles; are you more drawn to being a molder, caster or typographer in your craft/sculpting?

Also, it's always bugged me in testing going in and out of a custom games lobby to a forge lobby to test stuff; the party gets confused and it's jarring.

In my mind's eye, on this imaginary next-gen offering, I can see Matchmaking anything streamlined into one bucket, with your Campaign/Whatever OpsFight/PvP all available there, but switching from MM to Party for any one of these things "jumps" you to the appropriate lobby for Campaign or Whatever OpsFight(w/o randoms & more options) and Foundry, with your typographer, caster and core editors all there for all your custom pvp needs.

The UI that normally would show you in-depth info on your playlists, population, hoppers and game/map combos all in a good hierarchy, scale and easily accessible without digging but depth there if you want to dig deeper, would of course be replaced by all the knob/dial UI for those party areas where you have more options, while a DETAILED active roster is available ot the "front page" of EVERY lobby, again with appropriate scale, hierarchy and accessibility. The one place where you would have options available in the MM lobby would be your search preferences (options for DLC only, JIP on/off alongside your prefer good connection/close match/language option), player preferences, voice from/to randoms preferences.

Theater would of course ideally work with sessions from every play mode and be in its own lobby for however many people it could support, 4 would be amazing, 2 an acceptable target.

I dunno, the Foundry part seems pretty ambitious but for me at least map/gametype "shipments" for forgers and MM playlist designers, and the UI improvements would be the target to work toward in the future.

I'm gonna need a refresher. Why are certain ordnance drops labelled with Attacker or Defender? I was under the impression that they were supposed to be neutral.

What/where now? I have no idea, I picked apart Haven and Adrift's Initial/Random and never saw it, and AFAIK there is no reason to set Objective Ordnance to anything other than spawn order. What map are you seeing this on, a disc map or a custom?
 

Karl2177

Member
What/where now? I have no idea, I picked apart Haven and Adrift's Initial/Random and never saw it, and AFAIK there is no reason to set Objective Ordnance to anything other than spawn order. What map are you seeing this on, a disc map or a custom?

Haven. All of the initial ordnance items on it are tagged with either defender or attacker.
 

FyreWulff

Member
Could be left over from a dropped idea. Does a Defender trying to pick up an Attacker initial drop result in a pick-up meter (like when you try to pick up someone else's ordnance drop)?
 
[Graphics talk, skip if just here for forge :p]

I sent Corrinne Yu (343i Principle engine programmer [rasterization and lighting programmer]) a couple|of tweets regarding z-fighting in Forge, i'll post any reply here.

I think the z-fighting problem is a problem that could relatively easily fixed, if it gets attention. There's a few ways around it, but from my (limited) rasterization knowledge the best way (as mentioned in my tweets) to fix it would be to either increase the precision of the z-buffer values, and/or use a heuristic (or algorithmic) method to prevent the actual flicker between objects.

Another method would be to, upon map-file save, run a method that detects polygons that intersect and are parallel, and displace them by s small amount, such that when rasterizing, the depth values will never or rarely be equal, and z-fighting will be eliminated.

With the next gen xbox, hopefully increasing precision in the graphics pipeline will be much more plausible, but for arrays as large as depth buffers (currently 1280×720×16 bit? = 15.2MB of RAM), I hope it's possible.

Could be left over from a dropped idea. Does a Defender trying to pick up an Attacker initial drop result in a pick-up meter (like when you try to pick up someone else's ordnance drop)?

I haven't tested it, but I doubt it. If there is any change it would simply be the team which is marked would get a waypoint to the drop, whilst others would not. This would be pretty useful for symmetric maps.
 

Karl2177

Member
Could be left over from a dropped idea. Does a Defender trying to pick up an Attacker initial drop result in a pick-up meter (like when you try to pick up someone else's ordnance drop)?
No. The scattershot has the Attacker tag on it, but either team can pick it up. I had the thought that when it drops, it shows up on that teams HUD, but I set the blue team far from it and it didn't show up. Fairly certain that is all LoD based.
 
Top Bottom