Battle scale vs. graphic fidelity poll.

In regards to battle size scale (# of troops on screen) vs how detailed the quality of those troops

  • Larger scale battles with more troops on screen at cost of texture/model quality

    Votes: 41 53.2%
  • Smaller scale battles with high quality detail on clean crisp textures

    Votes: 5 6.5%
  • A mix somewhere in the middle

    Votes: 31 40.3%

  • Total voters
    77

Users who are viewing this thread

Hello folks, I once modded here, and as of late, Bannerlord has given me a itch to jump back in. I'm trying to ping the general population, and get an idea of what people's ideal experience would be when it comes to playing M&B.

Are you the type of person that would want to play battles, surrounded by hundreds or more troops sacrificing some texture quality and polygon count to push troop count to upper limits, or would you prefer smaller tight compact battles, where the troops would have greater detail, with the textures being best detail in first person, and equipment models having nice rounded edges... Or perhaps somewhere in the middle?

This feedback would help me in determining what's worthwhile now, and where I would want to lay my foundation for modding. I plan on a total conversion of some limited scale to start, and build from there.


Thinking about it now, I guess this could also be an argument of singleplayer vs multiplayer mode, and which is more prevelant nowadays. This too would shape my direction, so if you could as well post a response if you are more looking forward to singleplayer, multiplayer, or maybe both in your replies, that would be helpful.

Thanks for your time,

E.





 
ealabor said:
Hello folks, I once modded here, and as of late, Bannerlord has given me a itch to jump back in. I'm trying to ping the general population, and get an idea of what people's ideal experience would be when it comes to playing M&B.

Are you the type of person that would want to play battles, surrounded by hundreds or more troops sacrificing some texture quality and polygon count to push troop count to upper limits, or would you prefer smaller tight compact battles, where the troops would have greater detail, with the textures being best detail in first person, and equipment models having nice rounded edges... Or perhaps somewhere in the middle?

This feedback would help me in determining what's worthwhile now, and where I would want to lay my foundation for modding. I plan on a total conversion of some limited scale to start, and build from there.


Thinking about it now, I guess this could also be an argument of singleplayer vs multiplayer mode, and which is more prevelant nowadays. This too would shape my direction, so if you could as well post a response if you are more looking forward to singleplayer, multiplayer, or maybe both in your replies, that would be helpful.

Thanks for your time,

E.

8 k people who still plays warband is your answer. When the battle is greater, you have less time to look at details or graphics .
 
I’m interested in single player mods where battle sizes vary. However, I am looking forward to large scale battles and do not believe this will require a major downgrade in graphical quality on a current generation high-end pc.

Mount & Blade II: Bannerlord relies even more on CPU usage than Warband. Hundreds of characters, more advanced animations, an Inverse Kinematics system, individual AI, formation AI, combat calculations, (which do not change in respect to distance or visibility,) and many other requirements really increase the burden on the CPU. In order to accommodate this, our optimisation efforts are more heavily focused on the CPU. We generally try to use Data Oriented Design, which enables us to achieve high amounts of parallelism and core usage. Currently, 60-70% of the frame is fully parallel, which means it can, and will, use all of the cores of current and next gen CPUs for the foreseeable future, (the old engine generally used to use 1, or at most 2 cores.) This means that as new, higher core count CPUs begin to emerge, Bannerlord will scale well with the new hardware and players will be able to test bigger and denser battles. Currently our aim for battle sizes on current generation high end gaming CPUs is at 800 characters, at 60FPS.
https://steamcommunity.com/games/261550/announcements/detail/3244137880934579683

Average texture size is 2k. You can add 4K textures to the game.
https://steamcommunity.com/games/261550/announcements/detail/1464090667824385857

Polygon count for an armour set is usually between 8000-12000 triangles.
https://steamcommunity.com/games/261550/announcements/detail/1457340983216971711

IIRC they use 6 levels of LODs but one of them is for the campaign map so map icons wear the same armour as the characters.
 
There isn't a tradeoff between graphics and battle size in warband, and there probably won't be in bannerlord. The things that cause games with lots of agents to run slowly are mostly CPU actions while graphical fidelity (even if it could be quantified) is all on the GPU. In warband you could basically switch off the renderer by making the window tiny, or by looking straight at the ground to avoid rendering more than a few things, and it wouldn't really affect the framerate.
 
BIGGER Kentucky James XXL said:
There isn't a tradeoff between graphics and battle size in warband, and there probably won't be in bannerlord. The things that cause games with lots of agents to run slowly are mostly CPU actions while graphical fidelity (even if it could be quantified) is all on the GPU. In warband you could basically switch off the renderer by making the window tiny, or by looking straight at the ground to avoid rendering more than a few things, and it wouldn't really affect the framerate.
Been a while since I've played singleplayer but in multiplayer, depending on the map, that often isn't true.
Large maps with a lot of models will have a very noticeable framerate drop depending on where you look.
I believe that's because the game only culls models that are out of your field of view, not those that are out of your line of sight.
So on a large castle map for example, if you're standing outside the castle and looking towards it, framerate will drop.
Might be wrong on that though. Either way, looking at the ground will increase framerate in a lot of maps in multiplayer.
 
I should have clarified: Imagine you're playing a modded warband game where there are 1000 men each side. It would probably run at around 5fps because of the animations and calculations (which are CPU actions). Looking at the ground would make the GPU run a bit faster, but at the end of every frame it would still be waiting for the CPU to finish and there would be no visible improvement.

Of course if the GPU is the bottleneck, for instance when someone has made a terribly inefficient multiplayer map or your computer is old, then looking at the ground will increase the framerate because in this case the CPU is waiting for the GPU to finish at the end of each frame. This isn't often the case though, as warband was a fairly GPU unintensive game even back in 2010.
 
Ealabor, nice to see you around again; I was a keen follower of the progress of your mod for quite a long time. I just hope you don't have to go through as much work for Bannerlord!
 
The worst case are cluttered SP siege scenes with corpses all around the player, so lods are not used much. You still need to make sure that some potato laptops are able to run this too, or you lose some audience.

DanAngleland said:
Ealabor, nice to see you around again; I was a keen follower of the progress of your mod for quite a long time. I just hope you don't have to go through as much work for Bannerlord!
There's no Jhessail this time around, so let's hope he does make more progress.  :mrgreen:
 
Voted for option A.

I am only interested in singleplayer.

If there were a coop campaign multiplayer (where I could play with a friend or two) I would be interested in that, but definitely not the classic multiplayer.

In general I found multiplayer games, if it is not coop with close friends and nobody else, to be a very bad experience, because of an incredibly high rate of immersion killing children and/or trolls squatting on the servers.

I hope that helps.
 
BIGGER Kentucky James XXL said:
You'll have to deal with immersion killing children in SP as well by the looks of things.
What a coincidence. In my old Bannerlord mod plans, the game was supposed to start with a mass slaughter of children for maximum shock value (and instant immersion). You'll only hear the cries though because the perpetrators would put out the children camp torches at night. I'm not a monster like these Taleworlds fellows.
 
NPC99 said:
I’m interested in single player mods where battle sizes vary. However, I am looking forward to large scale battles and do not believe this will require a major downgrade in graphical quality on a current generation high-end pc.

Mount & Blade II: Bannerlord relies even more on CPU usage than Warband. Hundreds of characters, more advanced animations, an Inverse Kinematics system, individual AI, formation AI, combat calculations, (which do not change in respect to distance or visibility,) and many other requirements really increase the burden on the CPU. In order to accommodate this, our optimisation efforts are more heavily focused on the CPU. We generally try to use Data Oriented Design, which enables us to achieve high amounts of parallelism and core usage. Currently, 60-70% of the frame is fully parallel, which means it can, and will, use all of the cores of current and next gen CPUs for the foreseeable future, (the old engine generally used to use 1, or at most 2 cores.) This means that as new, higher core count CPUs begin to emerge, Bannerlord will scale well with the new hardware and players will be able to test bigger and denser battles. Currently our aim for battle sizes on current generation high end gaming CPUs is at 800 characters, at 60FPS.
https://steamcommunity.com/games/261550/announcements/detail/3244137880934579683

Average texture size is 2k. You can add 4K textures to the game.
https://steamcommunity.com/games/261550/announcements/detail/1464090667824385857

Polygon count for an armour set is usually between 8000-12000 triangles.
https://steamcommunity.com/games/261550/announcements/detail/1457340983216971711

IIRC they use 6 levels of LODs but one of them is for the campaign map so map icons wear the same armour as the characters.

Thank you, that's all very useful information. It even scratched a few itches I didn't know I had.

BIGGER Kentucky James XXL said:
There isn't a tradeoff between graphics and battle size in warband, and there probably won't be in bannerlord. The things that cause games with lots of agents to run slowly are mostly CPU actions while graphical fidelity (even if it could be quantified) is all on the GPU. In warband you could basically switch off the renderer by making the window tiny, or by looking straight at the ground to avoid rendering more than a few things, and it wouldn't really affect the framerate.

Both GPU and CPU processes were contributors to diminishing performance in regards to larger battles, in Warband (modded anyways), and i'm sure this would be the case for Bannerlord, as any other game for that matter. Although, reading into one of those links provided by NPC, I see that with Bannerlord, it would lean far more intensely into CPU than with the previous title, which is more to your point.

To make a short story long, with warband, as I scaled up the size battles it became a case of whack-a-mole between GPU and CPU, to keep ratcheting up the scale;

The first order of business for any sensible modder implementing custom meshes, is to make and implement lods' along with the custom meshes to take burden off GPU, which for me was always a priority. 268 B.C., in its latest iteration also had custom scripts for formations (had to have sarissa phalanx formations of course),  so I had to tweak some characteristics about the ai calls in the code, to then take burden off the CPU. After that it went back to GPU, in that as the battle size went up the reinforcement waves were causing a periodic stutter when they popped in, having to suddenly load lots of meshes and textures for a bunch of new troops in a largely still filled battle. I think the workaround I had for that it was to make more frequent, smaller wave sizes to absorb in without the periodic stutter.

The last hurdle I hopped was CPU. I kept getting instances where some battles run fine, and others wouldn't. I noticed that the slowdown in performance was happening occasionally on the random maps, but never on scenes I created. That was a head scratcher for a minute, but I had deduced that on the random maps it was happening on, they tended to be heavily forested. Thinking it was GPU related, I took a nerf bat to the tree meshes and textures, but didn't work. Then I decided to open scene editor on a random map, and realized yeah.. there's no AI mesh here. Pockets of troops were getting hung up on objects such as groups of trees and rocks, and going AI dead/AI crazy trying to process a path to their next enemy. So it then made sense why the big custom scene battles, where I spent hours making AI mesh to cover every nook and cranny were running much better than some of the random gen maps.

If the typical vanilla game size parties were around 60-120,  268 B.C.'s were up to 400-600 or more, last I left it. I can't remember what that translated to in regards to, in battle numbers.


DanAngleland said:
Ealabor, nice to see you around again; I was a keen follower of the progress of your mod for quite a long time.

Thank you Dan, nice to see you again too. Good to see there are still a few people around, keeping the memory of my work alive  :smile:

DanAngleland said:
I just hope you don't have to go through as much work for Bannerlord!

Well, I haven't created a mesh, opened python, or so much as painted a single pixel on a texture sheet since I stepped away.  Also, there are going to be a bunch of new systems to learn with Bannerlord... how much work could it possibly be?  :lol:

As far as mod scope goes, if I do commit to returning, its going to be something sensible that one person can handle workload wise, though whatever im working on i'd like to leave room for expansion. If there are any contributors along the way, that's well and good, though I don't think that I will be trying to create a small team this go around. Well, let me say if that does happen, joining members would be thoroughly vetted to avoid any... travesties, like with happened with the latest 268 B.C.

Rodrigo Ribaldo said:
The worst case are cluttered SP siege scenes with corpses all around the player, so lods are not used much. You still need to make sure that some potato laptops are able to run this too, or you lose some audience.

Wasn't there an option to turn the corpses off?

Rodrigo Ribaldo said:
There's no Jhessail this time around, so let's hope he does make more progress.  :mrgreen:

:lol:

Yep, even if she or any other of the few that I had lengthy debates with were still here, I wouldn't bite. I'm not compelled to spend any lengthy time in those off topic threads. I'm just here for the pixels this go around  :wink:

上原亜衣 said:
It appears they've gone for neither.

If in regards to the poll, I think its pretty telling that they have. Well at least we know they certainly aren't here for Assassins Creed Bannerlord, and more about the battle immersion. A median of the two highest choices gives a good sense of whats wanted, and where I, or any keen modder could spend poly count and texture quality, and where to save.

DerGreif said:
Voted for option A.

I am only interested in singleplayer.

If there were a coop campaign multiplayer (where I could play with a friend or two) I would be interested in that, but definitely not the classic multiplayer.

In general I found multiplayer games, if it is not coop with close friends and nobody else, to be a very bad experience, because of an incredibly high rate of immersion killing children and/or trolls squatting on the servers.

I hope that helps.

Yes it does, thanks for the response. I've always fiended for a coop since M&B first came out.
 
ealabor said:
The last hurdle I hopped was CPU. I kept getting instances where some battles run fine, and others wouldn't. I noticed that the slowdown in performance was happening occasionally on the random maps, but never on scenes I created. That was a head scratcher for a minute, but I had deduced that on the random maps it was happening on, they tended to be heavily forested. Thinking it was GPU related, I took a nerf bat to the tree meshes and textures, but didn't work. Then I decided to open scene editor on a random map, and realized yeah.. there's no AI mesh here. Pockets of troops were getting hung up on objects such as groups of trees and rocks, and going AI dead/AI crazy trying to process a path to their next enemy. So it then made sense why the big custom scene battles, where I spent hours making AI mesh to cover every nook and cranny were running much better than some of the random gen maps.

Taleworlds have abandoned randomly generated maps in Bannerlord. All their battle scenes are hand-crafted.

The size of scenes is generally around 4 km square, which is much larger than in Warband, and scene creation is much faster. We developed lots of new editing tools for objects, terrain and flora for the scene designers to use. We have an advanced terrain system which can support up to 16 layers per scene, with no restriction on layers used per node. Also, the designers now have access to a scene upgrade level system, (a scene masking system which is used in siege scenes to allow level 1-3 castles to be placed in one scene,) and a new weather system, (which is used for creating different weather versions of the scene.) With all of the CPU and hard drive optimisations we try to achieve <1 second loading times on these bigger scenes.
https://www.taleworlds.com/en/Games/Bannerlord/Blog/25

In videos, scene loading screens are visible for much longer than Murat Türe suggests.
 
NPC99 said:
Taleworlds have abandoned randomly generated maps in Bannerlord. All their battle scenes are hand-crafted.

Oh, that's unfortunate. AI pathing issues on gigantic scale battles aside, they were a great feature.

So I guess they have generated a pool of premade maps to be chose from at random, when having encounters on the map? Any idea how big the pool is?
 
ealabor said:
NPC99 said:
Taleworlds have abandoned randomly generated maps in Bannerlord. All their battle scenes are hand-crafted.

Oh, that's unfortunate. AI pathing issues on gigantic scale battles aside, they were a great feature.

So I guess they have generated a pool of premade maps to be chose from at random, when having encounters on the map? Any idea how big the pool is?

As far as I know, this is all they have said so far:

WHAT IMPROVEMENTS HAVE BEEN MADE TO RANDOMLY GENERATED BATTLE SCENES COMPARED TO WARBAND?
We don’t have randomly generated battle scenes in Bannerlord. With our new engine and new map editing tools, we are much more efficient at making terrains with no settlements on them. We realised that we can have enough handmade maps for battle scenes. We designated areas on the world map with biomes and tagged our battle scenes according to their features.

This helps us to have consistent quality and fun gameplay with all of our battle scenes, with hand placed spawn positions, managed distances according to party size and many other small details to spice up your gameplay.”

https://www.taleworlds.com/en/Games/Bannerlord/Blog/64

I suspect TW might have a revised their system of entry/spawn points to allow multiple start points on the same handcrafted map, allowing us to use each map edge, but I’ve seen nothing to confirm this. Despite the problems with mountainous terrain, a random battle generator keeps game replays fresh - however many battle maps TW have constructed, multiple games and mods will need more.
 
I wonder how many maps is enough variety for the long M&B campaigns. It needs to be a pretty large number.
It's fine there won't be extreme sinusoid maps anymore, but they needed that procedural generation.
 
NPC99 said:
As far as I know, this is all they have said so far:

WHAT IMPROVEMENTS HAVE BEEN MADE TO RANDOMLY GENERATED BATTLE SCENES COMPARED TO WARBAND?
We don’t have randomly generated battle scenes in Bannerlord. With our new engine and new map editing tools, we are much more efficient at making terrains with no settlements on them. We realised that we can have enough handmade maps for battle scenes. We designated areas on the world map with biomes and tagged our battle scenes according to their features.

This helps us to have consistent quality and fun gameplay with all of our battle scenes, with hand placed spawn positions, managed distances according to party size and many other small details to spice up your gameplay.”

https://www.taleworlds.com/en/Games/Bannerlord/Blog/64

I suspect TW might have a revised their system of entry/spawn points to allow multiple start points on the same handcrafted map, allowing us to use each map edge, but I’ve seen nothing to confirm this.

I'm uneasy about that decision. It has many benefits of course, such as maps looking more natural and with greater detail. But I am concerned about maps repeating, so I hope they have plenty of maps premade.
 
Back
Top Bottom