Not as large, but there's still significant gains in quite a few scrutinized titles as of late. And it seems to be often within 1-2 months of release. Usually preceded by a generic Twitter apology jpg.
Upscaling Technology Has Become A Crutch
- Thread starter BitByte
- Start date
Right, except remember the Nanite sampling level is biased towards the upsampled resolution, not the render resolution (similar to textures). Thus in Valley of the Ancients at 4k with 50% render resolution Nanite will be targeting ~1080p-sized geometry if my memory serves on the settings.
Some of these settings aren't technically hard to expose to users, but as I noted it would be pretty content dependent how much it affects performance and also how much wiggle room there is before obvious artifacts and popping would appear.
It has been said better before, but games are moving target, often right up until release. "How did this obvious bug or performance problem get through" is usually some combination of:
1) There were even bigger problems the week before
2) Content changed in the last week
3) A fix for last week's issue introduced this new one
The notion that the game was being played and tested in a similar state to the release for months or years in advance is usually not true. If there's a few weeks where content is locked and people get to focus on performance and stability that's pretty great. Obviously we would all prefer that more time get allocated to polish than is typical these days, and that's always good consumer feedback (don't buy it until it's acceptable to you). Sadly feature/content creep is very common as everyone really wants to get as much in as possible for release and it takes a large amount of discipline to gauge the right risk/benefit tradeoff as release approaches.
When I hear this, I always think of Dyack.
I remember after Too Human that he said the way the industry works is wrong and they should work on the title until its done, then announce it and launch it 6-8 months later giving even more time for bug fixes. This had to be like 20 years ago now maybe 15 ?
As great as it is to see such fast turn-around on these post-launch patches, it's not something you want to sit and reflect on because of its implications. Even if the performance gains are legitimate you're still looking at a case of something that could have shipped in a much better state but didn't. I have an easier time understanding an overly ambitious launch leading to months of big updates rather than an apathetic launch where fixes come quick after people complain.
I agree that some variation of that would be incredible. Instead of developers working down to the wire and putting in as much content as they can up until the release date.. they should be working down and putting enough content in as they can until the "announcement date".. When the game is announced, it's essentially content complete and from there the teams work on bug squashing, optimization, and polishing up until release.
Obviously there's going to be many legitimate reasons why it wouldn't work that way... or at least be very hard to change to that form of a development cycle for many devs/pubs.. but it sure would be nice. We've seen it time and time again where a game releases in a very rough state, and within a mere week or two after launch it's in a much better place than before. Developers should have more of a buffer than they do.
Was this an apathetic launch? How do people know it could have launched in a better state earlier? There were obviously some optimization options.
I think one thing people are missing is console releases do not have flexible dates. I think people assume that because some big games get delayed that every game can be delayed. It's probably not really an option for publishers and developers that have more limited finances.
Remnant 2 came out with heavy performance, but as far as I know it wasn't a broken game. It wasn't buggy and it had playable settings. It wasn't stuttery and didn't feel bad to play. They've optimized the performance within two weeks. It's less than ideal, but it's also not a travesty.
The best correction would probably be if idiots stopped pre-ordering games, and stopped blindly buying games on day one at full price. Companies know they'll get the sales even if they launch in a poor state, so they do. This is really a consumer problem. Not sure why people can't wait a week to see what's up, or wait for patches before buying.
You'd be looking at something like 6 month reduction in revenue. That is, games will be stalled an additional six months from now to get them ready for release. During those six months you won't be getting sales from these titles. Going forwards, you can overlap production so it's only this six month and it doesn't get worse, but that's a huge cost.
Is there any proven benefit? Have gamers shown they refuse to buy broken products and you need to release in a better state to secure income? Nope. So why would any publisher sacrifice that money? You'd need to show that that massive 6 month investment will pay big dividends down the line to please the board and shareholders into committing to it.
The overall development time would still be similar overall. You wouldn't be withholding a completed game for 6 months, you're essentially just announcing the game closer to it being finished and committing to a date when you have a clearer picture of what needs to be done to properly polish the game.
It feels like publishers lock their developers into dates too early, and devs have made a bunch of stuff that they don't want to cut or alter, and so they push and push to meet their visions, and the polish phase always pays the price for it. A lot of games seemingly need just a couple of weeks to a month of polish and optimization.. and it's amazing how many issues they can fix quickly after they've stopped trying to add/change things about the game.
But you're right that there's nothing forcing publishers to change the status quo... at least not at this point.
That's 6 months later than the current release dates.
Think about why we ended up in this state in the first place. There was nothing stopping pubs/devs holding off releasing later other than their own economic reasons. They could all choose to release 'when it's ready' and be three years longer. Or they could choose 'just another six more months'. Or they choose, "let's release it now anyway."
From a production POV: I don't like releasing anything unless I know it's pretty much done and bug free. I take 2x longer than everyone else to release anything! At least! The longer you spend on it, without a clear release date, the more you can see that you want to add and improve. Diablo 3 released in a very different state to that it was two years later. The devs could have held off for two years and made the superior "two years later" release the launch release. Doing so they would have made les money.
Alternative a dev can release simpler. Put in less and just make sure it's bug free. Now you've less of a game, less exciting and inspiring but hey, there's nothing that sucks about it.
On the flip side when you have something out there and users using it and liking it and complaining about it, you have something that's generating income and connections and motivation.
I could release my current project tomorrow. It'd be okay. But it's a bit flat, needs some more interest, needs more variety, different balance. So I'll work on it just a bit longer. Without a deadline, I can just keep working and working until it's perfect and the market has moved on.
The situation we have now is the natural product of free online patching and the difficulties of the creative process. You have to release at some point. Without the exacting deadlines of old-school Gold Master prints, there's no logistical way to nail down a game in development between ever-evolving features and bugs and consumers who will pay almost no matter what state it's in. Pubs aren't going to take on the costs without reasons. Until consumers give them the financial reason to adapt, they won't. They can't really, if operating their business as a business and not a hobby or charity. Business practices means maximising revenue means getting it out there when it hits that "this isn't too shit to sell if we market well enough" threshold.
Of course I know the reality is that even if you give developers an extra month... they'll use it to keep adding and making changes up to the last minute, and you'll still end up in the same situation anyway.. just a month laterThat's 6 months later than the current release dates.
Think about why we ended up in this state in the first place. There was nothing stopping pubs/devs holding off releasing later other than their own economic reasons. They could all choose to release 'when it's ready' and be three years longer. Or they could choose 'just another six more months'. Or they choose, "let's release it now anyway."
From a production POV: I don't like releasing anything unless I know it's pretty much done and bug free. I take 2x longer than everyone else to release anything! At least! The longer you spend on it, without a clear release date, the more you can see that you want to add and improve. Diablo 3 released in a very different state to that it was two years later. The devs could have held off for two years and made the superior "two years later" release the launch release. Doing so they would have made les money.
Alternative a dev can release simpler. Put in less and just make sure it's bug free. Now you've less of a game, less exciting and inspiring but hey, there's nothing that sucks about it.
On the flip side when you have something out there and users using it and liking it and complaining about it, you have something that's generating income and connections and motivation.
I could release my current project tomorrow. It'd be okay. But it's a bit flat, needs some more interest, needs more variety, different balance. So I'll work on it just a bit longer. Without a deadline, I can just keep working and working until it's perfect and the market has moved on.
The situation we have now is the natural product of free online patching and the difficulties of the creative process. You have to release at some point. Without the exacting deadlines of old-school Gold Master prints, there's no logistical way to nail down a game in development between ever-evolving features and bugs and consumers who will pay almost no matter what state it's in. Pubs aren't going to take on the costs without reasons. Until consumers give them the financial reason to adapt, they won't. They can't really, if operating their business as a business and not a hobby or charity. Business practices means maximising revenue means getting it out there when it hits that "this isn't too shit to sell if we market well enough" threshold.
Anyway, what I wish is for is all the bug fixing and optimization and patches that can happen in the 3 weeks after launch, to happen before the release date. Whatever discipline developers need to have to know when to quit building and start polishing, and whatever understanding publishers need to have.. I'd just like to see it happen. And I mean it does.. a lot. But it doesn't happen enough these days.
I'm generally quite understanding if the developers are very communicative of their intent to fix and improve things... but sometimes things are just bad and shouldn't be released until they're ready.
This is not yesteryear. Announcement well before release makes a lot of sense. Modern titles can build sizeable communities well before launch that keep gamers engaged and invested until the game is ready for retail.
Announcements act as a market tool and kicks off the initial excitement. Doing it further out gives the game plenty of time to build upon that existing enthusiasm.
The only problem with gaming is the state of titles at release. Creating a schedule with a realistic timeline, doing an honest assessment 3-6 months out, having a good QC/QA system and prioritizing a relatively bug free and stable release build is what pubs should be encouraging and help facilitating to and for their devs.
I think the Starfield delays, while disappointing, were actually well done. Coming months in advance of release dates or range and extending out well enough not to be a total rush job. Hopefully the game releases to minimal issues.
Announcements act as a market tool and kicks off the initial excitement. Doing it further out gives the game plenty of time to build upon that existing enthusiasm.
The only problem with gaming is the state of titles at release. Creating a schedule with a realistic timeline, doing an honest assessment 3-6 months out, having a good QC/QA system and prioritizing a relatively bug free and stable release build is what pubs should be encouraging and help facilitating to and for their devs.
I think the Starfield delays, while disappointing, were actually well done. Coming months in advance of release dates or range and extending out well enough not to be a total rush job. Hopefully the game releases to minimal issues.
Last edited:
We've certainly had a couple of high-tier titles get a couple of weeks last-minute clean-up delay. I guess no physical release helps.
You run the risk of a game being announced so early that people fall off the hype train. I remember when the KOTOR mmo was announced myself and all my friends were super into it but out of the group of 10 of us only 3 of us bought the game at release some 5 years after the announcement. People just fell off waiting on a game. You also have an isue when there is little known people fill in the gaps themselves and often times as the game solidifies it doesn't match what the people thought and move on.
Look at Gotham nights , it was announced way early , was delayed twice , last gen systems dropped from release and it still failed. If it was only announced 6 months before release they wouldn't have had the negative press of the dropped last gen consoles and the delays.
Yea, I imagine publishers want to know all that stuff as far in advance as possible and hold the teams to their estimations. They have milestones during production for a reason.. it helps them stay somewhat on track all throughout that time.
Here we go again with Immortals of Aveum. I might be wrong but from what I've seen, this game is basically a corridor shooter? 6800xt or 3080ti for 1440p/60fps using medium with upscaling? Excuse me but what? I mean, okay the game looks good but, not that good. Like only the 3090/ti, 4070ti, 4080 and 4090 are better than the 3080ti on Nvidia's side of the stack and that's what we're using for medium at 1440p with upscaling? I'm frankly not sure is just UE5 being UE5 but yea.... If it is down to UE5, i'm thoroughly unimpressed with it's performance and I hope developers don't abandon their technology in favor of this power virus(Hyperbole).
Last edited:
You have to move with the times. lifelike detailed modeling, Mesh Shaders and billions of micro polygons require such resources.Here we go again with Immortals of Aveum. I might be wrong but from what I've seen, this game is basically a corridor shooter? 6800xt or 3080ti for 1440p/60fps using medium with upscaling? Excuse me but what? I mean, okay the game looks good but, not that good. Like only the 3090/ti, 4070ti, 4080 and 4090 are better than the 3080ti on Nvidia's side of the stack and that's what we're using for medium at 1440p with upscaling? I'm frankly not sure is just UE5 being UE5 but yea.... If it is down to UE5, i'm thoroughly unimpressed with it's performance and I hope developers don't abandon their technology in favor of this power virus(Hyperbole).
However, don't forget that this game runs at 60fps on current consoles with Nanite lighting. If the TSR resolution and antialiasing are good enough, this game will be a good example of how UE5 looks good on today's consoles. A few days and it will be clear.
Most developer studios have closers whose role is to come in and cut items out and ensure that the game is in a shippable state.
The issue is likely that the game is close to the deadline and far from hitting it's actual game design goals and they are just left trying to complete items that should be well into the polishing stages. This is why we are seeing more day 1 patches and things that should be in the 'DLC'. These are just cut things that couldn't have made it and just pushed out.
Meanwhile, I see that this game uses FSR2 instead of TSR. But that doesn't change the essence of what I'm saying.
This happens every time there's a big change in rendering that has performance implications. There were people who thought Quake looked bad at the time because sprites looked cleaner. There were people who thought Virtua Fighter and Tekken looked worse than 2D fighters for the same reason. There were huge arguments of Half-Life 2 graphics vs Doom 3 graphics, because Half-Life 2 ran better and had sharper textures. Same with Painkiller at that time that ran very fast and had ultra sharp textures. This has always just been a thing. Some new tech comes along with big performance costs and some people are just going to prefer a low polygon tree with a nice bark texture slapped on top.