Sudden Strike 4 News / Updates Übersicht

[Last Soldier]

Seien wir realistisch, ich glaube auch nicht, dass die Unity-Engine aktuell über 4.000 Einheiten gleichzeitig darstellen kann (und, wie ich an anderer Stelle schrieb, bräuchte ich persönlich (!) auch nicht so viele).

Basically Unity can manage it if game objects and their logic, physics, timesteps etc. are out from standard Unity framework.
But certainly not on such a scale as if the units were 2D. If Sust can handle: 50000 map objects, 800 houses and 1000 units per player so new engine and PC today hw specs can handle maybe 10x more? In 3D its 10x less. Even though the view is limited by perspective camera.

Brauchen wir keineswegs. Es gibt mittlerweile Multithreading und DX12/Vulkan. Heutzutage sind sogar mehrere zehntausend Einheiten möglich. Das hängt davon ab, ob die Entwickler alles in nur einem Thread berechnen lassen wie RTS vor 10 Jahren oder vom Multithreading-Feature der Unity-Engine Gebrach machen. Für Sudden Strike wäre die Möglichkeit von tausenden Einheiten schon ausreichend für große Gefechte.

Ashes of Singularity is more AMD tech demo than standard game and they have direct support from AMD. If you look at screenshots so terrain has no details, there is 0 map objects, only few low poly trees and thats all + maybe 20 different units + game logic is very simple. All this helps render units in large quantities.

Btw. DX12 and Vulkan API's are big problem for large company, for small devs its dream. Maybe in the future. Unity has a generally poor performance - lags, drops, freezing, bigger games try to burn your computer especially if you have older graphics card and it's impossible to optimize on low level, almost everything is bottleneck. No threads or DX12 cant help there. Half of Sust2 is probably programmed directly in assembler a very well optimized so it makes a huge difference. Same Ashes of Singularity have own engine.

[ransombolton]

Seien wir realistisch, ich glaube auch nicht, dass die Unity-Engine aktuell über 4.000 Einheiten gleichzeitig darstellen kann (und, wie ich an anderer Stelle schrieb, bräuchte ich persönlich (!) auch nicht so viele).

Basically Unity can manage it if game objects and their logic, physics, timesteps etc. are out from standard Unity framework.
But certainly not on such a scale as if the units were 2D. If Sust can handle: 50000 map objects, 800 houses and 1000 units per player so new engine and PC today hw specs can handle maybe 10x more? In 3D its 10x less. Even though the view is limited by perspective camera.

Brauchen wir keineswegs. Es gibt mittlerweile Multithreading und DX12/Vulkan. Heutzutage sind sogar mehrere zehntausend Einheiten möglich. Das hängt davon ab, ob die Entwickler alles in nur einem Thread berechnen lassen wie RTS vor 10 Jahren oder vom Multithreading-Feature der Unity-Engine Gebrach machen. Für Sudden Strike wäre die Möglichkeit von tausenden Einheiten schon ausreichend für große Gefechte.

Ashes of Singularity is more AMD tech demo than standard game and they have direct support from AMD. If you look at screenshots so terrain has no details, there is 0 map objects, only few low poly trees and thats all + maybe 20 different units + game logic is very simple. All this helps render units in large quantities.

Btw. DX12 and Vulkan API's are big problem for large company, for small devs its dream. Maybe in the future. Unity has a generally poor performance - lags, drops, freezing, bigger games try to burn your computer especially if you have older graphics card and it's impossible to optimize on low level, almost everything is bottleneck. No threads or DX12 cant help there. Half of Sust2 is probably programmed directly in assembler a very well optimized so it makes a huge difference. Same Ashes of Singularity have own engine.

Actually, Last Soldier is right to some degree. I love Ashes of the Singularity and Stardock's work. We have close ties with them (again, even though you would think that they are some of our biggest rivals but love for strategy games kind of ties together). For example, we did the boxed versions for Sins of a Solar Empire, Galactic Civilizations and Offworld Trading Company. And Ashes is a greatly optimized game for what it does. It has an unrivalled number of units and a very large scale. BUT it does that at the expense of other things. Units are very basic and have a limited number of drawcalls (In Sudden Strike 4, each tanks has numerous drawcalls like turret, front, back, each effing wheel in the tracks, etc.). If I remember correctly, all of their units are hovering units, so there is no contact with the ground and thus, no impact (track marks, etc.). There are nearly no objects on the maps, no vegetation or cities or something (Wouldnt be good for Sudden Strike as a big part of the gameplay and atmosphere is the believable battlefield, the tanks impact on the locations, etc.). Also, effects (weapons, etc.) are kind of basic as well. Only lasers, etc. So no calculation of projectiles, armor, sloped armor, ricochets, etc. like in Sudden Strike 4. That brings calculations down to just the bare amount of units, which admittedly, the engine does very well. And Brad Wardell already said that Ashes will just be the starting point to see what they can do with the engine in the future. So I am looking forward to what they can do with it. But at the moment, it is not something to compare to Sudden Strike 4.
Honestly, you could do a lot with DX12/Vulcan/Metal. But in order to use it to its fullest potential, you would want to produce ONLY for those, leaving behind everything below that, everything that is DX11, OpenGL, etc. And that is not really a good alternative at the moment. Maybe in the future, but not now.

About Unity: Yes, it has its limitations. But not more or less than other engines when it comes to strategy games. Unreal, Cryengine are very focused on FPS or third person opitimization. For a while, Unity was focused on small-scale productions. And it was the first engine that opened up to small teams and was free to use and play around with. Naturally, that produced some good games and a lot of problematic results. That is just normal. And it somehow led people to believe that that was the fault of the engine when it really is not. Unity has evolved a lot in the last couple of years and I think it speaks volumes that all of Kalypso's currently announced big titles (Urban Empire, Vikings, Sudden Strike 4) are developed in Unity. Dungeons 2 was also developed in Unity. And yes, I think as a company, we are pushing the boundaries of what can be done for strategy titles with the Unity engine. As with every engine, there are things that you need to work around or do yourself. But it is a very versatile engine that allows rapid changes. And it is free for anybody to use so it would make even deeper modding endeavours possible. That was also something we thought about. As of now, we have not regretted our decision.

[kackspecht]

Dieses Zerpflücken von Postings ist eine Unart der Diskussionskultur in Foren, schon immer gewesen. Wenn du nicht in der Lage bist, deine Meinung zusammenhängend formulieren zu können, muss ich auf diese "Argumente" deinerseits auch nicht eingehen.

Danke für dein Verständnis.

Deine Ausrede wirkt als hättest du nur keine Argumente. Wenn dich das tatsächlich überfordert und du nicht in der Lage bist, darauf sachlich zu antworten, ist das nicht mein Problem.

Brauchen wir keineswegs. Es gibt mittlerweile Multithreading und DX12/Vulkan. Heutzutage sind sogar mehrere zehntausend Einheiten möglich. Das hängt davon ab, ob die Entwickler alles in nur einem Thread berechnen lassen wie RTS vor 10 Jahren oder vom Multithreading-Feature der Unity-Engine Gebrach machen. Für Sudden Strike wäre die Möglichkeit von tausenden Einheiten schon ausreichend für große Gefechte.

Ashes of Singularity is more AMD tech demo than standard game and they have direct support from AMD. If you look at screenshots so terrain has no details, there is 0 map objects, only few low poly trees and thats all + maybe 20 different units + game logic is very simple. All this helps render units in large quantities.

Multithreading and the low lvl API DX12(no draw call limitations like in DX11) are crucial to achieve tens of thousands of units. The speedup per additional core is almost linear.
If you don't believe it, try playing AoS on one or two cores. Old low poly single threaded RTS with much less details fail at so many units.

Here is some info on how AoS achieves this:
http://gamingbolt.com/ashes-of-the-sing ... ht-sources

AoS-Units are not low poly like they used to be in old RTS. Instead they have curved surfaces and (like the terrain) reflect lighting.
Here is a cutout from a screenshot at max zoom lvl:

aos 5k max settings.png (1.31 MiB) 8417 mal betrachtet

Btw. DX12 and Vulkan API's are big problem for large company, for small devs its dream. Maybe in the future. Unity has a generally poor performance - lags, drops, freezing, bigger games try to burn your computer especially if you have older graphics card and it's impossible to optimize on low level, almost everything is bottleneck. No threads or DX12 cant help there. Half of Sust2 is probably programmed directly in assembler a very well optimized so it makes a huge difference. Same Ashes of Singularity have own engine.

The AoS developers consist of only 50 Stardock and 6 Oxide Games employees. I can't see why Unity with its 1000+ employees would not be able to achieve a similar effective multithreading and DX12 implementation in thier engine.
Thier engine looks impressive. Like in pre-rendered scenes. I hope they will support Vulkan as soon as possible, too.

Only lasers, etc. So no calculation of projectiles

AoS has missiles with smoke trails and lighting from lasers and plasma bolts is being reflected by terrain and units. Also there is vegatation.
https://www.youtube.com/watch?v=ZepjUuYEjmo

Honestly, you could do a lot with DX12/Vulcan/Metal. But in order to use it to its fullest potential, you would want to produce ONLY for those, leaving behind everything below that, everything that is DX11, OpenGL, etc. And that is not really a good alternative at the moment.

That is why the AoS developers support both DX12 and DX11. Nobody demands a DX12 only game.

[Ingwio]

Moin Chris,
wenn Multiplayer bei Sudden Strike 4 so gut wird wie die folgenden Beispiele, dann bekommst du hier ein Denkmal

[Last Soldier]

Multithreading and the low lvl API DX12(no draw call limitations like in DX11) are crucial to achieve tens of thousands of units. The speedup per additional core is almost linear.
If you don't believe it, try playing AoS on one or two cores. Old low poly single threaded RTS with much less details fail at so many units.
AoS-Units are not low poly like they used to be in old RTS. Instead they have curved surfaces and (like the terrain) reflect lighting.

This is also crucial: http://i.imgur.com/f6GPoyh.jpg - scene with very low variability. Total War use some form of instancing too but with 1000x more details.
You can render tanks and infantry in same (curved) way? As you can see, it is tricky.

The AoS developers consist of only 50 Stardock and 6 Oxide Games employees. I can't see why Unity with its 1000+ employees would not be able to achieve a similar effective multithreading and DX12 implementation in thier engine.

+AMD. They will add it: https://unity3d.com/unity/roadmap but performance and universal engine? And you still need expert for new API in your team?
It just does not work together in my opinion, but maybe I'm wrong.

Personally, I do not need cool fancy RTS game. I'd rather have fair game, more units, great settings, normal view etc. I still have my own imagination http://www.marcofolio.net/images/storie ... _large.jpg

[ransombolton]

Multithreading and the low lvl API DX12(no draw call limitations like in DX11) are crucial to achieve tens of thousands of units. The speedup per additional core is almost linear.
If you don't believe it, try playing AoS on one or two cores. Old low poly single threaded RTS with much less details fail at so many units.

Here is some info on how AoS achieves this:
http://gamingbolt.com/ashes-of-the-sing ... ht-sources

AoS-Units are not low poly like they used to be in old RTS. Instead they have curved surfaces and (like the terrain) reflect lighting.
Here is a cutout from a screenshot at max zoom lvl:

Yes they are. And very nicely so. Curved surfaces and all. But that does not affect drawcalls actually. The units may have a curved, rather high-poly mesh but I does so by being one solid thing. There are for example no wheels on it that move individually, etc. if I am not mistaken. It is something that saves performance. Yet, it is still a technical feat to marvel at. Very nice.

Btw. DX12 and Vulkan API's are big problem for large company, for small devs its dream. Maybe in the future. Unity has a generally poor performance - lags, drops, freezing, bigger games try to burn your computer especially if you have older graphics card and it's impossible to optimize on low level, almost everything is bottleneck. No threads or DX12 cant help there. Half of Sust2 is probably programmed directly in assembler a very well optimized so it makes a huge difference. Same Ashes of Singularity have own engine.

The AoS developers consist of only 50 Stardock and 6 Oxide Games employees. I can't see why Unity with its 1000+ employees would not be able to achieve a similar effective multithreading and DX12 implementation in thier engine.
Thier engine looks impressive. Like in pre-rendered scenes. I hope they will support Vulkan as soon as possible, too.

Pretty easy explanation: Because Stardock/Oxide are creating their engine for a very specific kind of game on a very specific platform. Large-scale strategy games on PC. Whereas the Unity team needs to, as Last Soldier already said, create a universal engine. One that works for a multitude of game concepts and on nearly every platform (PC, Mac, Linux, PS4, PS3, Xbox 360, Xbox One, Android, iOS, Windows Phone, Browser, etc.). Hardcore strategy games are not a focus there. That is why I said we are kind of pushing the boundaries there. And it will help improve the engine itself as well. If you are only doing very similar games, it may make sense to put a lot of effort into your own game engine. Our Gaming Minds Studio has its own engine for trading/economic games. And it is great. But of course it needs constant improvements and it has grown over nearly 10 years. Paradox have a similar thing with their Clauswitz engine that is being used in Europa Universalis, Crusader Kings, etc. But if you ever played a long session of one of those games, you come to hate their loading times. Haha. And it is of course because that engine, although impressive, is aging as well.
To relaunch the Sudden Strike series, it made sense to use a versatile, all-around engine and not create our own one from scratch. This way, we could concentrate on creating a kick-ass game and not create platform integration, editors, etc. all ourselves. We have plans to create our own systems around it in the future. But I think to use Unity for Sudden Strike 4 was a good decision.

Only lasers, etc. So no calculation of projectiles

AoS has missiles with smoke trails and lighting from lasers and plasma bolts is being reflected by terrain and units. Also there is vegatation.
https://www.youtube.com/watch?v=ZepjUuYEjmo

Yes all the weapon effects are really nice. But if you look closely, they are just effects. They are not projectiles being calculated, etc. It is an effect that has effect on the surrounding (refleciton, etc.) and a clear effect on the target. Laser goes from a to be and does x damage. No probabilities or armor repercussion, etc. Also, yes there is vegetation, there are some trees and bushes. But so far, I do not see them having any effect on gameplay or being affected by units.

I am not saying it is worse. Ashes of the Singulariy is a great game. It just has a different focus and I am wary to compare it to Sudden Strike 4.

[ransombolton]

Moin Chris,
wenn Multiplayer bei Sudden Strike 4 so gut wird wie die folgenden Beispiele, dann bekommst du hier ein Denkmal

Wahnsinn. Habs mir gestern vorm Schlafengehen angeschaut. Also das Match zwischen Demon und ChubZ. Kam mir ein bisschen zu oft "freaking" drin vor aber ansonsten echt super Match. Und ich finde, es zeigt die Pros und Cons wunderbar. Die Mod, wie viele andere auch, erlaubt riesige Maps und Unmengen an Einheiten. Wunderbar. Damit kommen dann auch schöne Stellungskriege und Flankenmanöver, etc. zustande. Sehr schön anzusehen.
Aber: Mal ganz abgesehen von dem SuS-typischen Problemen (Einheiten sind kaum zu sehen, weil sie sich nicht vom Untergrund absetzen, etc.) gibt es hier auch einige Probleme. Ist es wirklich so toll und/oder realistisch, dass Gebäude so viel aushalten? Ich mein, Demon ballert bald eine halbe Stunde lang mit Panzern und Infanterie auf eine kleine Holzhütte und kann, weil die nicht down geht, den Punkt nicht einnehmen? ORLY? Und wenn ich dann mal so ganz kritisch sein darf: Was passiert mit den vielen Einheiten und dieser riesigen Map? Geschätzt über 65% der Map werden so gut wie gar nicht genutzt (ja, ich weiss, dass es eigentlich eine Vierspielerkarte ist) und es kommen immer wieder Verstärkungen, die sinnlos verheizt werden. Infanterie ist, außer zum besetzen von Gebäuden und scouting, so gut wie nutzlos und rennt den Panzern in die Arme. Und, wie Demon ja auch ständig im Video bemerkt, bricht das ganze herunter auf ein paar wichtige Einheiten. Seinen Crusader, die Pershings, ein paar seiner PAKs sowie auf der anderen Seite ChubZ Panther, Tiger, Königstiger und die 88er. Alles andere läuft so nebenher und wird schnell zu Schrott. Und im Grunde hat auch keiner der beiden die Möglichkeit, jede Einheit zu ihrem besten Potential einzusetzen, weil einfach zu viel gleichzeitig und an zu vielen Fronten passiert.

Now rage!

[Ingwio]

Also die beiden Videos sind ja auch nur 2 Beispiele. Mit Abstand ist der RWM 8.0 am besten und abgesehen davon gibt es von Mod zu Mod auch wieder riesige Unterschiede. Beim RWM wurde mehr auf Realismus geachtet und da sind die Häuser auch relativ schnell platt.

Auch wenn zum Teil 40 % der Map nicht genutzt werden, heisst das nicht das Spieler eventuell beim nächsten Match ganz andere Bereiche der Map nutzen.

Bei dem MP game hätte ich z.B. auch eine ganz andere Strategie gefahren. Ich flankiere immer gerne am Kartenrand, weil wenn man dort durchkommt geht es schnell voran, da viele das Hinterland meist sträflich vernachlässigen.
Es geht zum Teil einfach um die Möglichkeiten und ich glaube das sehen die meisten SS Fans hier genauso.

[Ingwio]

Hier dann nochmal ein RWM 8 Match. Also direkt im Spiel kann man die Einheiten eigentlich gut auseinander halten!

[Ente]

Hallo,

ob der Inhalt einer Karte genutzt wird, liegt

a) an der Kompetenz des Kartenerstellers
b) an der individuellen Spielweise des Spielers.

Wo soll der Vorteil sein, von vornherein die Kartengröße und damit die Intelligenz des Menschen vor der Bildschirm klein zu halten?

Mit der Einwandbearbeitungstechnik, den schwächsten Punkt des anderen zu treffen, um sich damit selbst positiv abzusetzen, so muss ich sagen, dass du, ransombolton, mit dem Holzhaus nun wirklich einen Aspekt angesprochen hast, den man individuell anpassen kann und der von Mod zu Mod verschieden ist.
Warum deswegen Sudden Strike 4 besser sein soll, erschließt sich mir nicht.

Gruß Ente

[Last Soldier]

What is going on behind the lines in MP?
- there are reinforcements routes, they must be secured (vs enemy paratroopers etc.)
- "secret" movements of army
- small parts of defense - AT and AA guns
- occupy cities, buldings
- lay minefields and build roadblocks
- artilery + amunition
- repair and oil trucks assembly point
- airport
- some players there holds most of its troops (but it's not very effective)

Battlefront is in RWM mostly 2 - 3 screens wide (1 screen = cca 800 tiles).

I don't know what is Ransombolton trying to say - in Sust4 with 10x lower number of units will be 90% of map empty? Haha, I understand the impassable forest now

Btw. if this is the maximal zoom so: (based on infantry size)

[Kongo-Müller]

Das mit der Sichtbarkeit der Einheiten finde ich extrem gut bei SuSt. Die wichtigen sind eh gruppiert und werden direkt ausgewählt. Es müsste halt möglich sein, mehr als zehn Gruppen zu bilden.
Bei großen Schlachten hätte ich gerne 100 GRuppen. Man kann sie nach Typ dann nummerieren.

Es ist klar, dass bei SS4 jetzt nicht mehr grundlegendes geändert werden kann. Ich werde es mir kaufen und spielen und dann hoffen, dass es ein Evolutionsschritt ist. Mit 3D wird es halt sowieso nie ganz das alte sein.

[kackspecht]

Multithreading and the low lvl API DX12(no draw call limitations like in DX11) are crucial to achieve tens of thousands of units. The speedup per additional core is almost linear.
If you don't believe it, try playing AoS on one or two cores. Old low poly single threaded RTS with much less details fail at so many units.
AoS-Units are not low poly like they used to be in old RTS. Instead they have curved surfaces and (like the terrain) reflect lighting.

This is also crucial: http://i.imgur.com/f6GPoyh.jpg - scene with very low variability. Total War use some form of instancing too but with 1000x more details.
You can render tanks and infantry in same (curved) way?

Of course you can. The Unity engine is even able to render units detailed like this.

The AoS developers consist of only 50 Stardock and 6 Oxide Games employees. I can't see why Unity with its 1000+ employees would not be able to achieve a similar effective multithreading and DX12 implementation in thier engine.

+AMD. They will add it: https://unity3d.com/unity/roadmap but performance and universal engine? And you still need expert for new API in your team?
It just does not work together in my opinion, but maybe I'm wrong.

Unity already supports DX12 and they "cannot wait to give" Vulkan "into the hands of our users".
Part of the support by Unity is to train the developers how to use thier tools and "Continuous updates with new features, platforms and improved performance."
Unity 5.4 supports "Fewer draw calls with GPU instancing" and "Better multithreaded rendering".

Pretty easy explanation: Because Stardock/Oxide are creating their engine for a very specific kind of game on a very specific platform. Large-scale strategy games on PC. Whereas the Unity team needs to, as Last Soldier already said, create a universal engine. One that works for a multitude of game concepts and on nearly every platform (PC, Mac, Linux, PS4, PS3, Xbox 360, Xbox One, Android, iOS, Windows Phone, Browser, etc.). Hardcore strategy games are not a focus there. That is why I said we are kind of pushing the boundaries there. And it will help improve the engine itself as well.

With DX12 or Vulkan the number of draw calls can be increased in Unity engine games, too. According to Unity they already increased performance by using DX12. This feature and multithreading is not limited to a specific engine. AoS is only an example for that. Another example is this benchmark with an increase of ~500% in the number of draw calls.

AoS has missiles with smoke trails and lighting from lasers and plasma bolts is being reflected by terrain and units. Also there is vegatation.
https://www.youtube.com/watch?v=ZepjUuYEjmo

Yes all the weapon effects are really nice. But if you look closely, they are just effects. They are not projectiles being calculated, etc. It is an effect that has effect on the surrounding (refleciton, etc.) and a clear effect on the target. Laser goes from a to be and does x damage.

Of course, in Sudden Strike 4 such effects are not necessary because usually in WW2 there are no fancy lasers and plasma projectiles flying around(except in alternative time lines in Star Trek ). When a WW2 tank gun is firing you only see the muzzle flash, a small smoke cloud and the impact. So there is no need to draw costly projectiles. Something similar like the current effect for MG projectiles in Sudden Strike 4 would be appropriate. Lighting effects would only be suitable for MG tracer rounds at night and smoke trails for missiles.

Currently Sudden Strike 4 tank gun and artillery projectiles and bombs cause a smoke effect like a rail gun from a FPS. Why?

No probabilities or armor repercussion, etc. Also, yes there is vegetation, there are some trees and bushes. But so far, I do not see them having any effect on gameplay or being affected by units.

That is not a reason to limit Sudden Strike 4 to a small number of units. Such calculations can be done in a separate thread and occur only at impact. Also WW2 tank guns are only firing at a slow rate.

I am not saying it is worse. Ashes of the Singulariy is a great game. It just has a different focus and I am wary to compare it to Sudden Strike 4.

AoS is an example on what is possible with multithreading and a modern low lvl API. Those are crucial to achieve tens of thousands of units. If you don't believe it, try playing AoS on one or two cores. The points is, there is no need to ignore todays technical possibilities to achieve at least thousands of units and limit the number of units to much less than known from Sudden Strike.

(Einheiten sind kaum zu sehen, weil sie sich nicht vom Untergrund absetzen, etc.)

Wer damit Schwierigkeiten hat, kann die Option "Units in Player Color" aktiviert lassen.

[Nebukadnezar]

Der Vorteil einer lizenzierten Engine ist immerhin, dass diese von einem anderen Team stets parallel weiterentwickelt wird.

@Chris: Gibt es einen engeren Austausch mit Unity Technologies? Könnt ihr mit denen über dieses oder jenes Engine-Feature sprechen? Oder habt ihr euch bei SuSt4 auf einer Versionsnummer festgelegt, so dass da keine Änderungen/Erweiterungen/Verbesserungen mehr zu erwarten sind, die man nicht selbst programmiert?

[Tyketto]

Was ich ziemlich schade finde ist das es keine Infos über die max. Anzahl an Einheiten oder die max. Mapgröße gibt. Das lässt sich doch ziemlich leicht testen und sollte zum jetzigen Zeitpunkt auch bekannt sein. Für mich ist das auf jeden fall eine Kaufentscheidung.