Do exclusive developers push visuals more than AAA developers?

Status
Not open for further replies.
Most third party publishers would call it a day and ship it like that (this is 3 months away from release)
(...)
If you have any example pointing towards the opposite feel free to post it.
What evidence do you have that above is at all true? I'd love to hear about a 3rd party developer that decided not to spend the last 3 months of development on polishing and tweaking.
 
Limited resources and budget have shown time and time again how some huge third party games were released where those 3 months of optimisation were clearly not enough.

The hard evidence is in the released games.

It's not about talent. There are no LazyDevs (Ok there must be some, but definitely not the majority). There is of course limited time and resources.

I would like to believe that most of us tend to do our jobs - whatever that may be - at the best of our abilities, but always within the constraints of our personal talent and determination, time and other resources, and with the influence of varying degrees of good/bad management from the high-ups.

Every job will be influenced by one of those. In my case in particular (nothing to do with developing games), my job seems to be influenced a lot by the decisions that are taken above my paygrade, and also time which never seems to be enough.
 
What evidence do you have that above is at all true?

Poor performance in most multiplatform releases? It seems as if most multiplatform games get released in a sub-optimal state, is it necessarily a problem of the hardware, or poor optimization? AC:Unity and Witcher 3 are prime examples of what i am talking about, games lingering the sub-30 fps range often enough for users to actually notice and complain about it. And both CDPR and Ubisoft promising patches that will fix performance in the future. So far as PS exclusives are concerned?

  • Driveclub: 1080p/30 fps capped 99.9% of the time, game looks gorgeous as well.
  • Second Son: 1080p/30+fps 99% of the time, small drops here and there when lots of particle effects appear on screen
  • First Light: 1080p/30+ fps 99.9% of the time, improved performance/visuals in comparison to Second Son
  • TLOU:R: 1080p/60 fps 99% of the time with fluctuations in some specific areas of the game

Could go on.
 
You don't know the causes of those framerate issues. Is it because the CPU is sitting idle 40% of the time as they don't have a suitable job system, or is the CPU idle 40% of the time because they have a job system but haven't optimised for it, or is the CPU 100% active and the framerate issue lies elsewhere? Your evidence is circumstantial and not proof of the alleged cause - that 3rd party devs won't optimise to the same degree.
 
You don't know the causes of those framerate issues. Is it because the CPU is sitting idle 40% of the time as they don't have a suitable job system, or is the CPU idle 40% of the time because they have a job system but haven't optimised for it, or is the CPU 100% active and the framerate issue lies elsewhere? Your evidence is circumstantial and not proof of the alleged cause - that 3rd party devs won't optimise to the same degree.

I know that i have no way of knowing why 3rd party games have such horrible performance issues most of the time. But the fact that such problems exist in the first place is a red flag pointing that something is going wrong in their workflow/engine w/e it is that 's bottlenecking the platforms. If the same case could be said for 1st party exclusives then it would be easy to call out the hardware as the limiting factor. It might just be that 3rd party developers don't have enough time to optimize their games to the same degree on both platforms and PC, that would still be a negative factor against 3rd party development when compared to 1st party development.
 
Poor performance in most multiplatform releases? It seems as if most multiplatform games get released in a sub-optimal state, is it necessarily a problem of the hardware, or poor optimization?
(...) Could go on.
This is not a proof of what you're claiming. There are dozens of issues - others than laziness - that cause suboptimal performance of various titles. It's clear to anyone who's ever written any production, multiplatform code or at least worked on something in a large-ish organization. Not to mention that your blanket statement can most likely be disproved anyway. But that's moot. The long story short: if you think that there's a single factor causing this, you're embarrassingly mistaken. Calling developers lazy is not a case of Occam's razor, it's a case of ignorance.
 
Poor performance in most multiplatform releases? It seems as if most multiplatform games get released in a sub-optimal state, is it necessarily a problem of the hardware, or poor optimization? AC:Unity and Witcher 3 are prime examples of what i am talking about, games lingering the sub-30 fps range often enough for users to actually notice and complain about it. And both CDPR and Ubisoft promising patches that will fix performance in the future. So far as PS exclusives are concerned?

  • Driveclub: 1080p/30 fps capped 99.9% of the time, game looks gorgeous as well.
  • Second Son: 1080p/30+fps 99% of the time, small drops here and there when lots of particle effects appear on screen
  • First Light: 1080p/30+ fps 99.9% of the time, improved performance/visuals in comparison to Second Son
  • TLOU:R: 1080p/60 fps 99% of the time with fluctuations in some specific areas of the game

Could go on.


You actually cited DriveClub which required an entire year after release to sort out all the missed features? This is hilarious set of blinders you must be wearing.
 
You actually cited DriveClub which required an entire year after release to sort out all the missed features? This is hilarious set of blinders you must be wearing.
And delayed nearly a year. DC is a good example of what most game companies will endure: hard timelines and limited budget. The difference is that with an exclusive you might get more budget and timeline when the product isn't there yet, with multi platforms the publisher can force you to ship, so the developers scramble by removing features and redesigning where they can to save time and money.
 
Yet another example is the release of Battlefield 4, Battlefield: Frontline and possibly Star Wars Battlefront at a 720p60 for X1 and 900p60 for Ps4. Three different game in the last two years in the same engine (Frostbite 3), yet no observable improvements in either performance or resolution. While Sucker Punch was able to improve their engine noticeably in just a five months (March 2014 Second Son, to August 2014 for first light)


You actually cited DriveClub which required an entire year after release to sort out all the missed features? This is hilarious set of blinders you must be wearing.
What do missing features have anything do with technical presentation exactly? The game was a mess at release, don't get me wrong, and i will never defend the state Driveclub was at that time, but performance was never an issue. It looks like you're trying to shift the argument where it suits you best :D
 
And this is how Driveclub performed day 1


If we talk about optimization of the hardware specifically (because this is what this thread is about) you can't fault Evo Studios, they provided a solid 30 at a native res of 1080p day 1. The game specific problems is another story, but i wouldn't say this is connected to optimization unless they optimized the game instead of having its features work which i don't think is true. I believe the Driveclub release is possibly the result of poor planning and setting unreachable goals for the team (PS+ edition alongside the full release), because you can't possibly blame what little time they had to release it, they got a year extension in the first place; in fact, DC was supposed to be a release title initially...
 
What do missing features have anything do with technical presentation exactly?
Reallocation of resources. "Okay boys, we can't get all this in there. All hands on deck to try and hit the intended framerate. Strip out anything slowing us down like weather effect. We'll optimise post release and reintroduce those stripped features when we can in patches.

AC could have launched with stable framerate if the design choices had been different and the devs had decided to remove content to make it run faster, then add that back as the engine is optimised. And optimising post release isn't necessarily good business practice, so it could be the AC devs are plenty capable of doing what ND or Insomniac do but they're governed by different rules and timelines. Give them another 3 months and perhaps the framerate would be solid and every bit the equal of your 1st party examples (whcih incidentally aren't always solid. I've seen my share of dropped frames and tears on Sony 1st party titles over the years too. Only Nintendo seems to really stick with the solid framerate mantra, although PS4 may have a different set of standards being adhered to).
 
There is no one game that does all. That's my point. I've seen several games that do something better than the other.

With regards to your observations on UC4.

1) The SSS on Nate in gameplay isn't very good looking. I've already mentioned this in another thread. But neither is The Order. I like the approach that CDPR used -- basically don't use it.

2) There is no SSS on hair. That's backlighting and it's used extensively on W3 for foliage as well as a myriad of other games. Remember that neither Nate nor most games use actual hair primitives while in gameplay. They are just strips of geometry with alpha maps on them. No AO either.

3) Deformable muscle -- again not in gameplay.

4) Explosions and smoke definitely look the best in Batman AK (at least the PC version w/Gameworks) and SW:Battlefront.

5) Fire? That's all artistic preference since all of them use sprites. I happen to think even the fire in the Order (especially the lamps) look incredible. It's also not surprising that one of the main artists for RAD now works at ND. She is extremely talented.

6) Physics -- I'll give UC4 that. But that's what they said they would concentrate on and they nailed it. However, it's not that they extracted some secret sauce of the PS4 that only it can do. Which is another one of my points. If any other dev wanted to focus on that, they could as well. As you said, Batman AK comes close just not to the same extent.

7) Foliage -- Ryse/Crysis 3/FC still has the best looking foliage to me. But the collison detection was also mastered by CDPR which used it extensively for pretty much all of it's movable foliage too. Again, not something magical that only ND has done.

8) Animation -- UC4 has excellent use of pre-baked animation. Looks really well done. But more impressive (from a technical level) is the procedural animation of the zombies in Dying Light. It's just incredible how they are unaware of obstacles and they fall, stagger, try to get up, slip on surfaces, etc..Which I think is much harder to implement in a game engine.

9) Textures -- clearly ND sacrificed texture resolution in that demo. While everything else seemed to come together nicely, the sheer texture detail was lacking. Again attributing to the limitations of the hardware -- not the engine.

Lastly, you didn't mention Lighting -- which is extremely important. The Tomorrow Children, Alien:Isolation, and Fable Legends all has taken GI a step further from baked light probes (which UC still uses). Batman AK has incredible amount of AO and shadow maps. W3, Dying Light, FC3, DriveClub all use dynamic shadowing from TOD. AC:Unity has the best baked GI maps for interiors we've ever seen.
1) No effect is perfect, all these AO, SSS hacks have flaws but it's better than nothing.
2) Still looks good but I guess you're right technically.
3) Yes it's in the gameplay, when he's stuck inside that flipped jeep. A default camera view is not most noticeable tho.
4) Sorry but neither Batman or SW got nothing on UC4's vehicle explosion, like by a wide margin.
5) If it looks good it looks good, fire certainly looks better in UC4 no regardless of sprites or fluid based. The Order does look comparable tho.
So you listed a whole bunch of games including TW3, The Order, Ryse, Batman etc to counter what UC4 one single game is doing, isn't that bit unfair? As I was saying earlier, UC4 has god tier balance, not only does it have a good variety of techniques but most of them are of good quality. Not a single game you listed has all those feats. Or another word, UC4 is the most well rounded game graphically speaking.
 
Sure, but poor frame-rate points towards the opposite.
No, it doesn't. I'm sure there's some sort of connection in your mind but there's none in real world. Unlike quite a few people on this board - you have very little experience writing code and working on large, cross-disciplinary projects. Please, educate yourself, then write from the point of authority. Unless you do that, I can pretty sure tell what's going to happen: people will stop discussing technical points with you on B3D and you'll eventually get tired of this place or you'll grow angry at people not getting your gospel. At which point you'll get banned. I'd strongly prefer that you'd actually listen and learn - there'd be more benefit for everyone involved if you would.
 
Reallocation of resources. "Okay boys, we can't get all this in there. All hands on deck to try and hit the intended framerate. Strip out anything slowing us down like weather effect. We'll optimise post release and reintroduce those stripped features when we can in patches.

But isn't that better than releasing a game like this?


I mean WHO even greenlit this? I mean, yes, Driveclub had it's faults, but it actually worked well enough as a game so that the player experience wasn't compromised, at least in single player.
 
No, it doesn't.

If poor framerate isn't connected with optimization, then does that mean that optimizing your rendering pipeline won't have an effect on framerate? I don't have a first hand experience on console game development but at least i know from first hand experience that optimizing your code can go a long way in improving performance. Maybe instead of writing lines about how your knowledge is superior, could you elaborate on what you are actually saying?
 
That's not the discussion. That's a business choice. For all you know, had DC launched with weather it'd have run at 15 fps. You can't use framerate as a litmus test for how optimised a game is, nor the reasons for those optimisations. And some would argue that AC without the crowds wouldn't provide the sense of populace needed, and it was the right choice.
 
That's not the discussion. That's a business choice. For all you know, had DC launched with weather it'd have run at 15 fps. You can't use framerate as a litmus test for how optimised a game is, nor the reasons for those optimisations. And some would argue that AC without the crowds wouldn't provide the sense of populace needed, and it was the right choice.

I agree that it's a business choice, as it is a business choice not to use expensive AA or AO techniques in your game to get better performance. My question is, why do multiplatform games tend to perform poorly in comparison to first party games day 1? It's surely a combination of factors, but do any of these actually excuse the state of the titles in question?
 
Status
Not open for further replies.
Back
Top