In contrast, ForzaTech continues to put forth some of the most visually and technically impressive racing games on the market.
Forza hasn't until now with Forza 7 (due to scorpio) been visually impressive, Forza 5 or 6 aren't visually impressive next to GT sport they are technically impressive because they push for framerate over graphical fidelity. Even if you look at last gen Forza compared to Gran Turismo, like F4 doesn't look visually impressive next to GT6. But a good example would be that Halo 5 is F4 visually not the best but a locked 60fps, were as GT6 is BF1 visually better but worse framerate (far worse in BF1's case.) Likewise Halo on base hardware can never hope to be visually best in class so long as it has technical goals and certain design goals, similarly Forza on base hardware can't hope to look better than drive club. But with the jump we have seen from Forza 6 to Forza 7 on scorpio, I would hope we would see a similar jump with Halo (fully fledged scalable PC version will help with that.)
Halo4 looked pretty good, and employed some nice effects, but the corners cut to achieve that look effected gameplay. Completely gone were the open sections and the sprawling vehicle sections. Ammunition disappeared before your eyes. It became a corridor shooter. The focus on 60 was the right call. That doesn't make the end result particularly appealing. When I speak of inefficiencies I'm talking about the ugly shaders and textures, the LODs, the animations, the fluctuating resolution, etc
"pretty good" from a purely technical point of view is a understatement, I think what needs to be pointed out is while Halo 4's version of the engine was on level with foundation, 4a and ND engine. The difference is that what are considered concessions with Halo 4 aren't concessions with a foundation, 4a or ND engine game, if Halo 4 wasn't called Halo these wouldn't be called concessions. It doesn't take away from the fact it is technically best of the best.
These are excuses. The number of game RELEASED doesn't mean anything. If anything 343s dev cycles should provide an advantage because they have more time to work with the hardware, especially as a first party dev. 343's locked 60 comes at the cost of a resolution that frequently bottoms out. The Xb1 version of BF1 spends most of its time near 60fps, has more players in a session, and has better textures, LODs, shaders and effects than Halo5.
These aren't excuses they are technical challenges, to call these excuses is a straw man. The number of games released has a clear effect on what the engine is capable of, that is even further shown due to the fact that the whole of EA uses and works on frostbite. Are you saying that for example ND's experience with LOU on PS4 didn't technically help them get their engine ready for UC4, or that their experience with UC4 hasn't allowed them to further improve with LOU2. You don't rip apart and rebuild different areas of your engine or implement different elements during mid or late development, so if you are dice and between each game you get to do so of course your engine will be further head. Given the fact they have multiple teams all working on the engine to improve it between multiple iterations and learnings compared to 343's one team, including the fact that EA have the equivalent of a studio dedicated to solely developing frostbite alongside every single one of their other studios doing the same. One of the reasons Halo 4 is easily the most technically advanced game on 360 is because they had a year or more before production even started to tear apart the Halo engine, bringing it up to the highest standard (which Halo had not been at since halo 1.) If you want to bring up Sony first party studios, well guess what they work just like frostbite were they are all managed by ICE team and each engine is upgraded and improved upon based on another Sony first party release (this is everything from a GOW lighting system being put in LOU2 or UC4 mo-cap tech being used in GOW.) The Turn10 team are unique in the fact that they share a live engine build between two studios who each year get to implement, rebuild and optimise the engine due to their release schedule. Thats what allows them to be at the forefront, but even then due to Turn10's technical goals they are slow to implement features. BF1 also definitely does not spend most of it time at 60fps, the game spends most of it's time at low 50's mid 40's dropping as low as the high 20's during the most intense scenarios. Players doesn't make a difference when you are dealing with a ton of AI, in terms of textures that is due to the benefits of photogrammetry. LOD is not better especially in multiplayer were there is a lot of pop-in, shaders and effects again are due to the benefits of photogrammetry. Nothing is taking away from the fact that BF1 during worst case scenarios is hitting 28fps at 720p, were as Halo 5 during similar worst case scenarios during warzone is still a locked 60fps. You can't expect Halo to visually be comparable to BF1 when they both have wildy different technical goals. If 343 let there framerates lowest threshold go as low as 28 fps they could give a equally as visually impressive game.
I think you just skipped over this point of my last post:
It's not simply double the framerate at worse case scenarios but that the difference in overhead between the two games at any time is large, BF1 during best case scenarios still doesn't hold a locked 60fps implying that there is little to no overhead. Were as Halo 5 at best and worst case scenarios holds a locked 60fps implying a large overhead. So not only is Halo 5 holding better performance in a similar environment but also while using less resources, that is efficiency.
A large overhead while also not sacrificing framerate means you have less resources you can put forward towards visuals, not to mention that BF1 sacrifices framerate as well as resolution.
I just don't understand this argument. Every single current gen developer had to transition to the next gen. My original statement was that I wish 343 had worked on a new engine INSTEAD of working on Halo4. Halo 5 would be Halo4 and would have had a normal dev cycle.
It's not an argument, the issue is you took one of the points out of the eight and chose to ignore the rest. A collective Point which is lost if you focus on one, "update the engine for current generation" isn't brought up as a point but as leveller. Updating an engine to current generation is already an undertaking in it's self for every other studio as you mentioned, it's the problem that all these other points on top of the rare event of a new generation that only happens every 6-10 years that further enhances the issues. What other current gen developer had to take an engine so intrinsically tied to 30fps, which had then been streamlined into a linear shooter and have it run at a locked 60 at all times (which I don't think you understand how hard of a technical goal that is) while also having it be capable enough to render a Halo 3 level of sandbox. Not to mention as they had serious development trouble as we now know (not related to the engine.) But the whole point of my "point" was that a new engine is a waste of money, time and resources in the end due to the very reasons I put forward.
I'd like to ditch dynamic res, add better texture filtering. Push the half-rate rendering much further back, fix whatever is weird about the dynamic shadows,better alpha effects, better LODs, return of over-the-top physics and dynamic AI, a better looking PBR solution, with enough overhead, with neccisary items being scalable to facilitate splitscreen and P2P
If you ditch dynamic res you also have to ditch a locked 60fps because the overhead isn't that large, or you make the overhead a lot larger in which case the graphical fidelity gets pulled back far further. Dynamic res is one of the best ways to use game consoles nowadays their is a reason why more and more developers are using it, 343's implementation is easily one of the most advanced. I do agree higher texture filtering at the cost of slightly lower dynamic resolution would be a good trade off, mostly due to texture filtering in Halo 5 having more of an effect on IQ than lower resolution. To push the half rate rendering further back you will need to free up some resources which would again further affect visuals, they could have figured this out for Halo 6 by pushing AI to the cloud or implementing a more dynamic version of the system. Rather than a hard lock 30 once you hit a static border you have a uncapped framerate that scale based on distance (ie 23yrds = 53fps/ 21yrds = 48fps.) There is nothing "weird" about the dynamic shadows they are there to avoid highly resource intensive scenarios, just like alpha effects and LOD which can all scale just as high as any other engine. It's the fact that there isn't the technical resources available to allow them, due to technical goals like a locked 60fps. "return of over-the-top physics" is something I would also like but is also a personal preference rather than a technical one, but nowadays they only have Halo fans left to cater to rather the majority of non-Halo fans who somehow hate how the warthog drives. "Dynamic AI" would ruin the Halo sandbox not to mention have a large performance hit, the Halo engine still uses Bungie's tree system. The issue with 343's AI is the lack of complexity in them especially the forerunners, and the fact that level design and the new sandbox elements have messed with the balance (also depending how a tree is designed it can result in AI appearing stupid or a lack/overload in difficulty.) PBR can definitely be improved, some areas look really good like Sanghelio's while others look like plastic. They have already confirmed splitscreen and P2P would be a massive step backwards, they are better of letting you pick the server you want to run of and improving the coverage of their servers. If they can't do that then depending on the distance from the server it switches to P2P. "with enough overhead" I don't think this is the right term for what you are trying to describe, because if it is it doesn't make any sense.