Update #34: FIRST ART UPDATE

[Ignatius]

Most excellent. 10/10 would read again.

 

Love the wizard. He feels like a hedgewizard, perhaps part of an ambush party v. party battle somewhere in the wilderness. I can see it.

Edited December 5, 2012 by Ignatius

[Kimozabi]

I really like being shown how the characters might look from the isometric gameplay perspective. And nice with a preview of a level layout as well.

 

Looking forward to this game like mad.

[Gyges]

I definitely like the little combat avatar model. Looks to have more detail, less awkward, better than the old stuff already.

Thanks. Glad you like it. The model in the scene is the same model in the image above (with different textures and a different weapon, obviously). We want the level of detail in the game models to be high enough that we can use the same model for the inventory screen "paper dolls" with a slightly magnified view.

 

Hmmm.. while you're here.... Will there be a chance to pick your brains with a pneumatic drill anytime soon? Is anyone planning to interview you in a similar fashion?

[Rob Nesler]

this game will be rendered in parallel projection. things there stay of the same size regardless of where in space they are. there is no horizon line*.

 

*or rather - it's in the infinity - *ba dum tsss*

And it's not a problem.

 

I was just curious about it when he mentioned rendered realism, and because you can see the deliberate perspective in their concept art.

*And that since the characters are [presumably] fully 3D, and not pre-rendered, it is possible to scale them down (in realtime) as they walk away from the viewer.

 

(However... I can see the potential problem when panning the view.)

the properties of the camera that is used to render the background are provided to the game engine and matched perfectly. There is no perspective. This in-game camera will be fixed on a general target location around the center of the party and dolly with them as they move around the world on an invisible collision plane that is coincident with the visible terrain. the in-game camera will never rotate around this target nor is it intended to zoom in or out.

[Rob Nesler]

Thanks for the update, was a great read!

 

I'm curious though, is the game intended to be viewed at a fixed perspective as with infinity engine games or will the camera orbit?

 

I'm looking at that tower model and noticing all the doors are accessible from "this side" of the model. I was just wondering if that's because this is the only side that will be visible or if this was the more interesting side to show.

There will be no perspective. All cameras are orthographic. The cameras and the world are aligned and essentially fixed at a specific distance, wich ww will iterate on, but will be pretty close to the experience that one would have had in Icewind Dale rather than Torment. The world image is static and pre-rendered. The camera will not rotate around to see the back-side of a building.

[PrimeJunta]

No zoom, then. Huh.

[Rob Nesler]

That was a sample of a larger map section. The viewscreen would be a much smaller window panning around that image. We are expecting to render maps that are up to 20k pixels wide, if the loading performance permits it.

I don't know about lefties, but I would like to develop different rigs for male and female characters.

 

Just out of curiosity, how does this compare to other Obsidian games like the Fallout series, ToEE, etc.

Those games had backgrounds that were rendered for VGA displays (640x480) at the low end of screen support we are quadrupling those numbers and targeting a 6x9 aspect ratio, at the high-end we might be approaching 10 times the pixel count. IWD maps were around: 3840x2880 pixels

[Jarmo]

I'm probably pretty much alone with this, but I'm a bit disappointed the character models are so high polygon.

Means we probably won't be seeing really big mobs of enemies, tribe of goblins having 100 warriors coming at you at once (instead of coming in 5 waves of 20),

or fight a group of 75 mercenaries.

 

Which would have been something I'd have liked.

I was a bit disappointed when NWN2 didn't improve on this compared to NWN (about the same amount of creatures seen at once on both, only in NWN2 with higher detail.)

(and way disappointed in Skyrims fight for the city battles where armies were about 10 guys on each side)

 

 

Oh well, it's not like I wouldn't appreciate good looks anyway.

[Ieo]

the properties of the camera that is used to render the background are provided to the game engine and matched perfectly. There is no perspective. This in-game camera will be fixed on a general target location around the center of the party and dolly with them as they move around the world on an invisible collision plane that is coincident with the visible terrain. the in-game camera will never rotate around this target nor is it intended to zoom in or out.

 

Okay, a mod needs to reword that sticky in General for "known information" about camera zoom. I personally don't care either way, but some may really want it.

 

One of the biggest annoyances in DA:O for me was the forced camera shifting--on a selected party member, unable to move the camera view far away into the fog of war or to the fringe of a given battle. At least that I remember, the camera in the old IE games would move with your party as you dropped those movement carats (no idea what they're called) on the ground and directed party location, but you could easily move the camera view manually to an unexplored part of the map or around an entire combat zone. I think there was also an option to center the camera view on a character during an auto-pause event. I hope PE will have similar, perhaps even greater, player-camera freedom.

[C2B]

The Neogaf Thread on the subject has become unberable and crossed lines.

 

Does someone have an account over there?

[Falkon Swiftblade]

Jarmo brings up something I was wondering. It would be great if theres a good mix of small swarms of things, maybe bats or sprites, as well as big bison/elephant sized creatures mixed in. Similar to how more recent games like Reckoning had wolves and big tubby Ettins (forgot the name) in the same fight so the player had fast attacks and big whammy attacks in the same fight making combat more strategic. It's important every once in a while to have spots mages can open the whammy and take out a room of baddies. One thing I hope we can avoid is only having like 8 enemy's in the same game over and over. I'm hoping for 30-50 unique ones at least. I think Skyrim had ~ 50 including all the humanoids which felt a lil samey even though it was a big world to explore. Though I guess it depends on how much combat is going to be the focus. If it's gonna be fewer fights I hope they're really tough.

[Tychoxi]

that building (first pic) looks great, great, great. I must say, though, that I'm expecting different architectural styles for different areas from you fine and revered game designers. And the characters models look just like what I was expecting (I don't mean the character's design but the level of detail and such), so it's looking good to me!

 

now, what had me going waaaa and WTF and wheeee was the "+3 medicine" flail, so I'm glad you guys removed Mark's blacksmithing privileges!

Edited December 5, 2012 by Tychoxi

[Rob Nesler]

I'm probably pretty much alone with this, but I'm a bit disappointed the character models are so high polygon.

Means we probably won't be seeing really big mobs of enemies, tribe of goblins having 100 warriors coming at you at once (instead of coming in 5 waves of 20),

or fight a group of 75 mercenaries.

 

I encourage our character artists to start high on the poly count. Edair is under 4k--so "high" for this kind of game. This allows for more flexibility, as we can always LOD the assets down to lesser numbers, as needed--in some cases, automagically.

The number of enemies will not be restricted because of poly count. intelligent combat AI, pathing and collision avoidance are the more costly CPU processes that will determine enemy counts, not to mention quite simply: "Designer Requirments" for an entertaining and manageable battle.

[Frisk]

We want the level of detail in the game models to be high enough that we can use the same model for the inventory screen "paper dolls" with a slightly magnified view.

 

OK, this answers one question from the Quality of 3D models.. thread from the tech forum

 

1) Have different LOD models for game and inventory screen (drawback: extra work)

 

2) Use only high-LOD models (drawback: unnecessary heavy in real-time)

 

3) Use only lower-LOD models (drawback: Inventory screen might look "ugly")

 

 

They are going for option #2 - meaning that the in-game models will really be a higher quality than is actually needed (or rather, higher level of detail than the graphics card is really capable of displaying, which will mean a little extra work for the hardware, but gives us better quality. I'm not complaining - that's for sure.

 

And then they are talking about cloth physics ... I wonder if that will mean we are going to get cloaks that look realistic with no clipping?

[Frisk]

I encourage our character artists to start high on the poly count. Edair is under 4k--so "high" for this kind of game.

 

Hmmm... higher than I had expected - I had estimated 2K in some post in some other thread, but again, I am glad to hear that you are not sacrificing quality. And, as you said, you can alwais scale it down easily if it becomes necessary.

[Ieo]

I'm probably pretty much alone with this, but I'm a bit disappointed the character models are so high polygon.

Means we probably won't be seeing really big mobs of enemies, tribe of goblins having 100 warriors coming at you at once (instead of coming in 5 waves of 20),

or fight a group of 75 mercenaries.

 

I encourage our character artists to start high on the poly count. Edair is under 4k--so "high" for this kind of game. This allows for more flexibility, as we can always LOD the assets down to lesser numbers, as needed--in some cases, automagically.

The number of enemies will not be restricted because of poly count. intelligent combat AI, pathing and collision avoidance are the more costly CPU processes that will determine enemy counts, not to mention quite simply: "Designer Requirments" for an entertaining and manageable battle.

 

I'm not a fan of swarm enemies (DA:O being a good example of few enemy types but lots of them), but flexible combinations offer more diversity in combat, sure.

 

Also, about this graphic density/quality per others' concerns---this is a basic rule for art in general, back when I worked with physical->digital graphics. Always start out with as high a quality as you can manage for the initial base assets, then scale down for functionality, even if the "product for the masses" ends up significantly compressed or something. Thing is, as hardware power increases over the years, this approach definitely has much better "aging." I don't see this approach as wasteful (higher than is needed) but rather future-proofing. Assuming that's what Obsidian has in mind too.

[rjshae]

Do we know that the inventory character images will have the same poly count as the in-game character renderings? They may also be using much lower poly count for in-game creatures.

[Frisk]

Do we know that the inventory character images will have the same poly count as the in-game character renderings? They may also be using much lower poly count for in-game creatures.

 

That's what he said:

 

We want the level of detail in the game models to be high enough that we can use the same model for the inventory screen "paper dolls" with a slightly magnified view.

[AwesomeOcelot]

I would have guessed 2K for that character, but then that's not taking into account animation, it looks more than fine at the perspective and distance the game camera will be. Not that poly count matters for that much, Dragon Age: Origins characters apparently use 10K-12K polygons each, which is higher than Half-Life 2's 6K-8K, but you can make up your own mind about that. It's not how large a poly count you have, it's how you use them, and by that I mean have great texture artists that work with the various maps because that's clearly more important in terms of how good characters look. Some of Assassins Creed 2's models use less than Half-Life 2's models, and they look pretty good. Animation trumps all of that anyway, it's way more important to get the animation right.

[rjshae]

^^^^

That and shadows.

[Tsumoro]

Really looking forward to more of these updates about the art, it's going to be so great to experience a new, living, breathing world. Because the character dolls are quite high in textures I am hoping that this will allow for some beautiful looking and unique weapons and armour in the game. I'm all one for half plate and full plate, but I do like those 'rare' finds when you find yourself with something that just looks simply amazing.

[Grotesque]

as a civil engineer, all I can say is that the doors and windows on the house are a little too big, disproportionate.

[ORDUCKMANIAC]

Cool update guys! I love the medicine ball lol! It got my creative mind spinning... Why not make a weapon that looks like a medicine ball and then when you get a critical hit spikes fly out of it for an AOE... Just thinking out loud... )

[anek]

The model in the scene is the same model in the image above (with different textures and a different weapon, obviously). We want the level of detail in the game models to be high enough that we can use the same model for the inventory screen "paper dolls" with a slightly magnified view.

 

With character models this detailed, faces will be clearly discernible (unlike, say, in the Infinity Engine games).

 

What will this mean for the protagonist? Since, according to earlier kickstarter updates, the player is supposed to be able to freely conceive pretty much everything about the protagonist they want to play as, from personal background to career to age to alignment/beliefs, this should be reflected in his or her appearance, and especially the face.

 

In the Infinity Engine games this was easy - character models were so small and pixelated that not much could be discerned anyways, so a character's look was pretty much defined by their portrait - which was just a 2D image that players could easily customize.

 

How will Project Eternity solve this?

Will the character creation screen give users so many customization options for the protagonist's 3D character model, that one could, say, create a 50 year old fighter with a rough and scarred face and steely glance? Or a 19 year old mage with a soft, tri-angular face, narrow eyes, and an appeal of playfulness and irresponsibility?

 

Or will we simply get a few fixed 3D models to choose from?

[AwesomeOcelot]

as a civil engineer, all I can say is that the doors and windows on the house are a little too big, disproportionate.

 

Could be because they have to accommodate people of unusually large size.