Hello Responders,

Earlier this week we launched the first of our two game health deep dive blogposts, which focused on bugs, how we triage between different severities of bugs, and the work QA does to deliver the best quality of game for our players.

This blogpost focus is on the work our programmers are doing to optimize the game to run at a high quality on as many rigs as possible. The EA launch version of No More Room in Hell 2 featured beautiful environments, realistic animations, and high level visual effects that really captivated players, but at a high technical cost. High level rigs were able to get the full immersive experience (most of the time) while a large number of players were stuck trying to play the game while dealing with a roller coaster FPS ride that sometimes crashed, killing responders.

The game has come a long way from its EA launch version, but there’s a lot more opportunity for improvement before we reach our 1.0 candidate.

Game performance is measured in how many frames a player sees in a second (FPS), and everything on a player's screen contributes to the weight of a frame in milliseconds (one thousandth of a second) and microseconds (one millionth of a second). If we’re looking to get a stable 60FPS each frame needs to happen in 16.667 milliseconds.

All of the individual assets, animations, and vfx that show up on a player's screen have a millisecond (or microsecond) cost that impacts the time it takes for your computer to process a frame, and leads to the number of frames that you see in one second.

Strong game performance is a difficult goal to achieve when you’re a smaller studio with big aspirations. Our artists and designers work much closer to the game build, and generally priorities a high quality threshold over game performance.

Our performance specialist programmers have been using new tools, ingenuity, and the growing strength of Unreal Engine 5 to further optimize performance while mitigating the sacrifice of quality.

Game performance becomes exponentially more difficult with each different each different computer's hardware combination, and almost every computer is a little different. We run our performance tests on multiple rigs representing minimum specs, recommended specs, and a high powered unit, but that still leaves a lot of room for players' unique computers to impact their individual performance.

Performance problems can sometimes be obvious, but sometimes they're not as obvious as well. Things that may seem fine in a vacuum can be problematic at scale; when many characters are active, performance issues can compound - even microseconds can matter! Some of the optimization work we've done recently may seem fairly minor, but when applied to many characters at once can actually lead to significant savings. We have to be careful though when making certain changes as there's a delicate balance between performance and visual quality. It's easy to say "just do things less often", but that's not always the answer if it causes unwanted behavior or choppy animation & movement. Sometimes we have to get a bit creative, which means it may take longer than expected to fix problems.

Our next update contains optimizations in many different areas for both CPU & GPU that should lead to noticeable framerate improvements for most players: character animation, shadowing, asset rendering, and more. Optimization continues to be a focus for us, so look out for even more in future updates!

– essbuh - Tech Lead

Optimization has been happening behind the scenes update after update since the launch of early access, but improvements might have gone under the radar, as they’ve come slowly. Over the next 3 months we have a few tricks left up our sleeves to optimize players' experience.

The current in-game building (left) has 250+ individual meshes. Through the use of our automated process we’re able to compress the same building (right) to just over 53 meshes, an 81% reduction!

Our team has been running a new automated optimization tool that looks at unique environmental assets and compresses as many meshes together as possible. The quality of the assets shouldn’t go down, but your computer will have less meshes to render, giving a big boost to performance. We’ve finished our pass on Lewiston and Broadway, and players will be able to experience the changes with our upcoming September update. We’re currently working on bringing the same updates to Powerplant and Pottsville in the near future.

Another example of our automated process at work, the building above went from 301 meshes to 24, for a 92% reduction.

Unreal Engine 5 is an extremely powerful engine that’s still relatively young, making each UE5 update a potential powerful tool for game optimization. Unreal released its latest update in June with 5.6, and we’re almost ready to make the leap. While it may take multiple updates for us to fully take advantage of the new optimization opportunities, the update should further optimize players' experience.

If you find yourself experiencing giant FPS drops during intense moments of the game there are some quick win setting adjustments that you can make:

• Shadows 

• Motion Blur

• Bloom

• Lens Flare

• Depth of Field

If you experience FPS drops that feel off for your rig and usual gameplay experience a Customer Support ticket[NMRIH2.com] is always helpful.

A reminder of what to add in your tickets:

• As much context as possible

  • What happened?

  • With who?

  • What map?

  • What location of the map?

  • Any other pieces of information that help further flesh out the situation you found yourself in

• A description of your rig and technical situation

  • GPU and CPU

  • Ping and the region you’re playing in

  • ~FPS

• Video or photo (if possible, they help a lot!)

If you want to join the conversation, hop into our Discord, leave feedback in-game, or visit

https://NMRIH2.com/support.

Hello Responders,

Last week we revealed that 1.0 will be coming in 2026, and we also touched upon a production pivot that has us focusing more on perfecting the current in-game features and offering players an optimized and polished gameplay experience.

September will bring a health focused update, similar to our Reinforcement update last November where we highly prioritize the health of the game over adding new features to NMRiH2. Game health is a mixture of optimization — the art of delivering the highest quality game in the smallest possible package — and polishing of the gameplay experience by removing bugs and confusing or unintended issues.

This first deep dive is all about our QA team, the work they’ve put into the game over the past 10 months to get the game to where it is today, and our ambitious polishing objectives for 1.0 next year. Every Support ticket[nmrih2.com] and crash report is a key component to game health, and offers a tremendous amount of help to our relatively small QA Team.

Every Quality Assurance (QA) team at every development studio works differently, but they all generally have the same mission: supporting the team in delivering the best possible product to its players. At Torn Banner, our QA team wears many hats, which include but are not limited to:

• Finding as many bugs as possible in as many iterations of each upcoming feature and update as possible.

• Troubleshooting those bugs and finding the root cause, then giving as much information as possible to the correct developer to fix the issue.

• Working with our Community team to track customer support tickets, feedback on platforms with Twitter, Reddit, and Discord,  and triage and troubleshoot issues as they come up.

• Entrench themselves within different development groups and support feature development as it happens

• Support and run internal and external playtests every week

• Generate game analytics reports & QA reports

“We have a lot to do as a QA team at Torn Banner, which can be a perk of the job! For me, being a liaison to every different department and developer at the studio is why I enjoy games testing so much. It also means our team has a responsibility to find bugs and feedback in each area of the game; in one day we might need to hunt down an objective bug blocking our next map, places you can get stuck in our live maps, and test a new weapon, new customizations, and new progression features. It’s a lot to juggle!”

—TornSheep, QA Tester

Torn Banner is a small studio, with a relatively small QA team, and that makes managing a large list of tasks difficult, No More Room in Hell 2 being an 8 player game makes everything more complicated, because so much testing has to be done in groups. We cannot be everywhere at once, and that makes triaging issues the most important part of the job.

We have to take a few things into account:

• How bad is a bug if it happens (total blocker, there are workarounds, just a blemish)

• How often a bug happens (every single game, some players might see it, you really have to look for it)

• Balancing our time between live bugs and bugs in upcoming features

With all of the factors above bugs are categorized as either minor, moderate, or blocker/major severity. 

“Severity is our shorthand to assess: “how bad is this bug going to be for a player?” As a QA team it’s important for us to be in alignment and test the most severe potential problems first, and it’s also key for us to communicate each bug’s Severity to the Production team and other departments, to understand the impact that any bug is having on the experience.”

—TornSheep, QA Tester

Figuring out the cause of every bug can be an extremely difficult and time consuming task. Sometimes it takes a lot of investigation, or just plain dumb luck, to recreate a bug that a player has seen and get the data we need to fix it.

Critical Objective bug:

For a while now players have been running into a severe issue where players on Pottsville were unable to interact with a transformer that was critical for opening up the next segment of the map. Through tedious trial and error the team eventually figured out that if you collected the objectives under CRC Responder body 2 & 3 at the exact same time it would cause the bug.

When testing maps, the stereotype is often "Time to run into this wall 300 times to see if it breaks!" - and while this absolutely is a part of what we do, we also have to look at objectives, replication (ensuring the game state is shown equally to all players on a server), and more subjective QA work like loot and zombie spawn balance.  The biggest new content we add to the game are our maps, and they get developed and iterated on over the course of many months.  Because of their size, there are *hundreds* of map-specific bugs that get found and fixed, but some bugs are rare, or introduced late in the development cycle.  Unfortunately, no studio in the world can match the sheer number of in-game hours that thousands of live users will put in on a content launch day.  There's always one missed hole in the wall, place someone can get stuck, or rare misbehaving objective.

— TornPotassium, QA Tester

This is a prime example of the type of “meat and potatoes” investigation and problem solving QA does on a daily basis, and it also shows how seemingly simple issues can be extremely difficult to solve. This specific bug has a fix that will go live in our September health update.

“Almost every single bug report — whether that comes from someone inside the QA team in our morning smoke test, from a developer in a weekly full-team playtest, or from a player reaching out on Discord or our ticketing system — gets a QA ticket for investigation. Sometimes these problems are easy to find and ping a developer to fix, but other times when we need server logs and performance profiles and the steps to recreate the bug are so rare, it can be a real collaborative effort to solve the problem. It can be like feeling around a dark room for a light switch.”

—TornSheep, QA Tester

Hit Registration:

Another really important, and complicated issue that our QA and Programmers are constantly dealing with is hit registration. A lot of independent variables go into hit registration, but the biggest is ping and servers.

To try and get a deeper understanding of the issue our QA team started the tenuous task of testing each and every weapon in various settings to gain a deeper understanding of how they all interact with the game and our servers. Through this, we’ve already fine tuned multiple factors related to combat and hit registration (we hope you’ve noticed!) but we’re not done yet.

At the end of the day, your ping causes latency, and that latency causes hit registration “mis-hits”. We’re hoping to be able to further refine and update the way our servers interact with players to trim down players ping and lower the amount of hit registration issues.

Our QA team’s primary goal still remains finding critical bugs and helping developers to remove them from the game, but we’re also starting to put a stronger emphasis on “polish” bugs, those that take away from players immersion but don’t necessarily impact the core gameplay loop of the game. 

Often in the gaming community there can be a sentiment that bugs get into the live game because QA testers ‘aren’t doing their jobs.’ But our testing team has a lot of ground to cover every day, and we have to make trade-offs about which bugs to investigate and log. Knowing about as many bugs as possible is just the first step. Helping developers and producers understand the most important fixes to make first is the other half of the battle. On top of it all, it’s very hard to predict exactly how complicated and time consuming a given bugfix can be! A lot of hands touch every single bug – from Community team members, to QA, to Production, to Developers, and back to QA again.

—TornSheep, QA Tester

The “polishing” and focus on our game health works hand in hand with our goals to prioritize getting the current features in No More Room in Hell 2 just right, and offering players a fully immersive, engaging, and replayable 1.0 experience.

During this Early Access period your help with tickets[nmrih2.com] and technical support feedback is more important than ever. While our priority needs to remain triaging tickets and targeting critical game breaking issues, we want to use this newly earned time to drop the overall “sitting” ticket number, and polish No More Room in Hell 2 until it’s in a place where it truly shines.

A reminder of what to add in your tickets:

• As much context as possible

  • What happened?

  • With who?

  • What map?

  • What location of the map?

  • Any other pieces of information that help further flesh out the situation you found yourself in

• A description of your rig and technical situation

  • GPU and CPU

  • Ping and the region you’re playing in

  • ~FPS

• Video or photo – if possible, they help a lot! – especially videos of things you did leading up to the bug, not just a video of the result!

If you want to join the conversation, hop into our Discord, leave feedback in-game, or visit

https://NMRIH2.com/support

.