71 replies to this topic

[Quicksilver Aberration]

Jin Ma, on 21 February 2015 - 08:36 PM, said:

I don't see how they would resize the texture without losing fidelity. They might remap it. But thats probably at least as much work as doing the texture in the first place

Regardless of that, scaling a mech and its rig doesn't require it to be re-textured. So that's neither here nor there.

[Nik Reaper]

Well, at leas it does seem certin that the executioner will have the advantage of the nipple hardpoints, so the question is also how high will he be able to point them up and down ( and if he can make the propellers turn X) ) because the arms seem to be gargoyle high, still sure hope that the mech is not higher than an atlas.

Edited by Nik Reaper, 21 February 2015 - 10:57 PM.

[Tennex]

I think if it follows the trend where the legs are shared. Then it will be our tallest mech yet



Eh maybe they can fix it through a more stooped animation or something

Edited by Tennex, 22 February 2015 - 04:36 AM.

[xengk]

cSand, on 20 February 2015 - 09:19 PM, said:

Look at the executioner's face

Spoiler

more like this

Spoiler

[xengk]

Jin Ma, on 21 February 2015 - 03:51 PM, said:

Will pay good money for a windup toy mech like the old Zoids toys.

[Tennex]

xengk, on 22 February 2015 - 05:46 AM, said:

Will pay good money for a windup toy mech like the old Zoids toys.

if zoids is old then i dont want to be young

[Helaton]

Jin Ma, on 21 February 2015 - 08:36 PM, said:

I don't see how they would resize the texture without losing fidelity. They might remap it. But thats probably at least as much work as doing the texture in the first place

Just want to comment on these. I'm not familiar with CryEngine, but familiar with Maya, Blender, 3DS. It is generally correct that it is simple to scale an object to whatever size you want prior to rigging. But depending on the object and complexity of rigging, it does not mean that you can scale an object post-rigging (its a case by case basis - could be possible).

Now think of the components that have to fit each hard point. The AC/UAC, MG, LAS, PLAS, PPC, GAUSS, etc variations of mounts that would be loaded on the mech. To keep continuity, an AC-20 shouldn't look like an AC-5 on a Shadowhawk compared to an Atlas. Regardless of the scale, some components cannot be scaled. What if the Grid Iron's hunch couldn't fit the AC-20 or Gauss after rescaling? It would have to be remodeled so it could fit (more fine tuning model adjustments which takes time.)

Texturing is a different beast altogether. Textures are typically flat mapped images applied to the surface. So take the example of the catapult example. If you scale it down by 22% and scale the texture down by 22% it will most of the time line up. (You could even leave the texture at full resolution anyways and let the renderer scale it down, but let's assume you scaled it down for the sake of optimizing the texture for size/performance.)

Also note that textures that go from large to small can preserve quality, while going from small to large will blur the details of the art depending on how much its enlarged.

The other example of the Centurion is different because its proportion is different. You usually can't adjust a texture that is out of proportion by stretching/condensing it to fit without the texture itself looking stretched/condensed. In this case, it would be better to start the texture over and salvage what you can.

After rigging & textures, the next problem is integration into the game engine. Whether scaled or altered, collision mechanics would need to be adjusted, hit boxes adjusted, hard point locations on the model adjusted. You probably can't just scale these down. Think if you have an AC20 coming out of that Centurion and it is scaled down to the size of AC-5, it doesn't keep proportion with the rest of the game.

Then there's time spent on game balancing (holy, its hard to hit the new catapult because its so tiny compared to before, but it carries those dual gauss, new meta etc. etc.) so the total time involved is more than just scaling a few models.

I would imagine that each mech for these changes and testing is probably around 70+ hours per mech scaled and then tweak time. Let's say over 6-12 months, another 50-70 hours from all parties involved. It adds up very quickly, and they probably don't make much money on redoing old content as they do releasing new content.

If its not broken enough, don't fix it. As a community, best bet would be to put together a kick starter to contribute money to go to PGI to make it worthwhile for them to fix the mechs while concentrating on the bottom line. All the people who agreed to the change, put a little money where their desires are. Just realize to change existing mechs takes time (potentially unprofitable time for PGI), and would probably need $30-50k of TLC to get up to date and make us happy.

That's probably the best truth of the situation I can think of.

[Revis Volek]

Praetor Knight, on 20 February 2015 - 09:18 PM, said:

I'm curious to see how those back panels, behind the head, are handled for hitboxes.

I imagine it will all be CT, Maybe if we get lucky they will cut the fins in half and let them straddle the CT and ST hit boxes. Maybe a 50/50 split when looking head on. Other wise it will take all DMG in the CT all the time. Even if all they see is its ears sticking up.

I hope it looks something like this....(sorry the pic is so junky, was rushing to get it up here at work)


The green blip in the center along with what i did not outline as the ST would all be CT and should end up being a decent Hitbox.

Edited by DarthRevis, 24 February 2015 - 02:54 PM.

[TELEFORCE]

Regarding the size of shared components between Clan Omnimechs: I doubt the Clans would waste resources in resizing similar limbs just because they're put on heavier or lighter chassis.

Just because the chassis weigh differently doesn't mean that the size must be different too. The Mad Dog is roughly the same size as the Timber Wolf, but the structure is built lighter since it is a 60-ton omnimech and not a 75 tonner. Therefore, the Mad Dog is the same size, but not the same weight, despite sharing the same leg and pelvis castings with the Timber Wolf. The same can be said with all the other Clan omnimechs.

[Tennex]

Helaton, on 24 February 2015 - 02:19 PM, said:

Just want to comment on these. I'm not familiar with CryEngine, but familiar with Maya, Blender, 3DS. It is generally correct that it is simple to scale an object to whatever size you want prior to rigging. But depending on the object and complexity of rigging, it does not mean that you can scale an object post-rigging (its a case by case basis - could be possible).

Now think of the components that have to fit each hard point. The AC/UAC, MG, LAS, PLAS, PPC, GAUSS, etc variations of mounts that would be loaded on the mech. To keep continuity, an AC-20 shouldn't look like an AC-5 on a Shadowhawk compared to an Atlas. Regardless of the scale, some components cannot be scaled. What if the Grid Iron's hunch couldn't fit the AC-20 or Gauss after rescaling? It would have to be remodeled so it could fit (more fine tuning model adjustments which takes time.)

Texturing is a different beast altogether. Textures are typically flat mapped images applied to the surface. So take the example of the catapult example. If you scale it down by 22% and scale the texture down by 22% it will most of the time line up. (You could even leave the texture at full resolution anyways and let the renderer scale it down, but let's assume you scaled it down for the sake of optimizing the texture for size/performance.)

Also note that textures that go from large to small can preserve quality, while going from small to large will blur the details of the art depending on how much its enlarged.

The other example of the Centurion is different because its proportion is different. You usually can't adjust a texture that is out of proportion by stretching/condensing it to fit without the texture itself looking stretched/condensed. In this case, it would be better to start the texture over and salvage what you can.

After rigging & textures, the next problem is integration into the game engine. Whether scaled or altered, collision mechanics would need to be adjusted, hit boxes adjusted, hard point locations on the model adjusted. You probably can't just scale these down. Think if you have an AC20 coming out of that Centurion and it is scaled down to the size of AC-5, it doesn't keep proportion with the rest of the game.

Then there's time spent on game balancing (holy, its hard to hit the new catapult because its so tiny compared to before, but it carries those dual gauss, new meta etc. etc.) so the total time involved is more than just scaling a few models.

I would imagine that each mech for these changes and testing is probably around 70+ hours per mech scaled and then tweak time. Let's say over 6-12 months, another 50-70 hours from all parties involved. It adds up very quickly, and they probably don't make much money on redoing old content as they do releasing new content.

If its not broken enough, don't fix it. As a community, best bet would be to put together a kick starter to contribute money to go to PGI to make it worthwhile for them to fix the mechs while concentrating on the bottom line. All the people who agreed to the change, put a little money where their desires are. Just realize to change existing mechs takes time (potentially unprofitable time for PGI), and would probably need $30-50k of TLC to get up to date and make us happy.

That's probably the best truth of the situation I can think of.

Thanks for the knowledge. Its nice to know the obstacles PGI has. For us to be informed consumers/complainers

But regardless of the work I still think its very important to scale these mechs correctly. Personally I wouldn't mind skipping 1 new mech if they could spend that time resizing 2 of the old mechs.

Of course i have no way of saying that with my wallet. But If there was a way, I would vote with my wallet.

[Quicksilver Aberration]

Tennex, on 24 February 2015 - 04:39 PM, said:

Thanks for the knowledge. Its nice to know the obstacles PGI has. For us to be informed consumers/complainers

But regardless of the work I still think its very important to scale these mechs correctly. Personally I wouldn't mind skipping 1 new mech if they could spend that time resizing 2 of the old mechs.

Of course i have no way of saying that with my wallet. But If there was a way, I would vote with my wallet.

Some of those are misleading as the texutre would not have to be modified. From Atlas to Lolcust, all textures have a preset size that they use. He is right that scaling the mech would affect how it looks on the mech but considering the mech textures are 1024x1024 and scaling would most likely not exceed 90% of the original size, it won't be significantly noticeable.

The rig itself can also be scaled without worry, but the animations may not translate as well so this is where time would be spent, ensuring the animations can carry over to the scaled rig.

Hitboxes should be able to be simply scaled as well as I know with old outdated MW4, hitboxes were always more simple version of the model and were a separate file.

All in all, a particular mech in MW4 (the Ares) was scaled up because of its size in comparison to other in translation to another mod, and between 3 developers, the process did not take that long, so I cannot imagine it taking that much longer in this game. Granted there is a huge difference between the engines of old and CryEngine, but I still can't imagine half of what you said is actually necessary.

The most time consuming of this process would be testing to make sure everything works out alright.

[Helaton]

I would bet textures work a little different and might either be simpler, or more complex depending on how it was handled with the ability to swap out primary/secondary/tertiary colors and patterns. It probably has a base texture and then the individual colors masked onto the pattern. Would be interesting to see a composite of the different textures going on the model.